This PDF file contains the front matter associated with SPIE Proceedings Volume 8856 including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

This paper presents a two-layered lossy-to-lossless still image coding system for medical images. The proposed system is compliant with widely used JPEG standard, in the sense that the base layer of the system is compressed by a conventional JPEG encoder. The enhancement layer includes extra information of the difference between the original medical image and a JPEG compressed version. It is compressed losslessly and its bitstream is embedded in the base layer codestream. Performance evaluations of single-layer and two-layer lossless compression are both carried out on a set of medical images. Single-layer lossless compression results show that JPEG 2000 delivers similar performance as JPEG-LS, and that both outperform JPEG XR and PNG. The experimental results of two-layer lossless compression demonstrate that the proposed system outperforms an earlier proposed JPEG based coding system with residual coding, and yields similar results as JPEG 2000 and JPEG-LS standard while providing quality scalability and backward compatibility to JPEG.

Telescope images of astronomical objects and man-made satellites are frequently characterized by high dynamic range and low SNR. We consider the problem of how to enhance these images, with the aim of making them visually useful rather than radiometrically accurate. Standard contrast and histogram adjustment tends to strongly amplify noise in dark regions of the image. Sophisticated techniques have been developed to address this problem in the context of natural scenes. However, these techniques often misbehave when confronted with low-SNR scenes that are also mostly empty space. We compare two classes of algorithms: contrast-limited adaptive histogram equalization, which achieves spatial localization via a tiling of the image, and gradient-domain techniques, which perform localized contrast adjustment by non-linearly remapping the gradient of the image in a content-dependent manner. We extend these to include a priori knowledge of SNR and the processing (e.g. deconvolution) that was applied in the preparation of the image. The methods will be illustrated with images of satellites from a ground-based telescope.

Image fusion methods that can effectively combine information from different imaging sources have become more common due to the growing types and number of imaging sensors, and because it can be difficult for humans to effectively integrate visual information by viewing multiple displays separately. In this paper, we show a novel techniques that runs at the first stage of the fusion process. The algorithm determines the "polarity" of the images (i.e., assigning brighter or darker gray-level values to larger image intensities), under the assumption that fusion produces better results when the images are more similar. The proposed technique was implemented to various fusion methods, including arithmetic fusion, Lapalce pyramids, Ratio of Low Pass pyramids and Morphological pyramids. Pairs of thermal and visual images were employed in the fusion procedure. The results indicate that the proposed enhancement leads to a better image fusion quality, with low computational cost, according to both objective and subjective fusion evaluations.

Correlation filters for object recognition represent an attractive alternative to feature based methods. These filters are usually synthesized as a combination of several training templates. These templates are commonly chosen in an ad-hoc manner by the designer, therefore, there is no guarantee that the best set of templates is chosen. In this work, we propose a new approach for the design of composite correlation filters using a multi-objective evolutionary algorithm in conjunction with a variable length coding technique. Given a vast search space of feasible templates, the algorithm finds a subset that allows the construction of a filter with an optimized performance in terms of several performance metrics. The resultant filter is capable of recognizing geometrically distorted versions of a target in high cluttering and noisy conditions. Computer simulation results obtained with the proposed approach are presented and discussed in terms of several performance metrics. These results are also compared to those obtained with existing correlation filters.

Cubic-spline interpolation (CSI) scheme is known to be designed to resample the discrete image data based on the leastsquares method with the cubic convolution interpolation (CCI) function. It is superior in performance to other interpolation functions for digital image processing. In this paper, an improved CSI scheme that combines the leastsquares method with an eight-point cubic interpolation kernel is developed in order to improve the performance of the original CSI scheme. Either the FFT/Winograd DFT or the fast direct computation algorithm can also be used to perform the circular convolution needed in this improved CSI scheme. Furthermore, its correlated image data and auto-correlated filter coefficients are also accurately calculated in this paper. Experimental results indicate that the proposed improved CSI scheme yields a much better quality of reconstructed image than existing interpolation algorithms.

Most of the classical approaches to the measurement and modeling of electro-optical imaging systems rely on the principles of linearity and space invariance (LSI). In our previous research efforts we have focused on measurement and analysis of images obtained from a double station video observation system MAIA (Meteor Automatic Imager and Analyzer). The video acquisition module of this system contains wide-field input lens which contributes to space-variability of the imaging system. For a precise astronomical measurement over the entire field of view, it is very important to comprehend how the characteristics of the imaging system can affect astrometric and photometric outputs. This paper presents an analysis of how the space-variance of the imaging system can affect precision of astrometric and photometric results. This analysis is based on image data acquired in laboratory experiments and astronomical observations with the wide-field system. Methods for efficient calibration of this system to obtain precise astrometric and photometric measurements are also proposed.

function of two variables distorted by a known linear operator and additive noise is usually used in the problem of image restoration. It was shown that to solve the problem metrical as well as topological characteristics of the function of two variables should be used. In paper with the help of spectral analysis we derive conditions when the gradient decent method is able to restore images. Computer simulation results are provided to illustrate the performance of the gradient decent method for restoration of uniformly blurred signals.

We propose a rate-efficient, feature-agnostic approach for encoding image features for cloud-based nearest neighbor search. We extract quantized random projections of the image features under consideration, transmit these to the cloud server, and perform matching in the space of the quantized projections. The advantage of this approach is that, once the underlying feature extraction algorithm is chosen for maximum discriminability and retrieval performance (e.g., SIFT, or eigen-features), the random projections guarantee a rate-efficient representation and fast server-based matching with negligible loss in accuracy. Using the Johnson-Lindenstrauss Lemma, we show that pair-wise distances between the underlying feature vectors are preserved in the corresponding quantized embeddings. We report experimental results of image retrieval on two image databases with different feature spaces; one using SIFT features and one using face features extracted using a variant of the Viola-Jones face recognition algorithm. For both feature spaces, quantized embeddings enable accurate image retrieval combined with improved bit-rate efficiency and speed of matching, when compared with the underlying feature spaces.

Image denoising manages to recover a digital image from its noisy version by exploring the statistical features inside a given noisy image. Most denoising methods perform well at low noise levels but lose efficiency at higher ones. In this paper, we propose a novel image denoising method, which restores an image by exploiting the correlations between the noisy image and the images retrieved from the cloud. Given a noisy image, we first retrieve relevant images based on feature-level similarity. These images are then geometrically aligned to the noisy image to increase global statistical correlation. Using the aligned images as references, we propose recovering the image with patch-level noise removal. For each noisy patch, we first retrieve similar patches from the references and stack these patches (including the noisy one) into a three dimensional (3D) group. We then obtain the noise free (NF) patches by collaborative filtering over the 3D groups. These recovered NF patches are aggregated together, producing the desired NF image. Experimental results demonstrate that our scheme achieves significantly better results compared to state-of-the-art methods in terms of both objective and subjective qualities.

In modern society, many companies offer product customization services to their customers. There are two major issues in providing customized products. First, product manufacturers need to effectively present their products to the customers who may be located in any geographical area. Second, customers need to be able to provide their feedbacks on the product in real-time. However, the traditional presentation approaches cannot effectively convey sufficient information for the product or efficiently adjust product design according to customers’ real-time feedbacks. In order to address these issues, we propose vPresent , a cloud based 3D virtual presentation environment, in this paper. In vPresent, the product expert can show the 3D virtual product to the remote customers and dynamically customize the product based on customers’ feedbacks, while customers can provide their opinions in real time when they are viewing a vivid 3D visualization of the product. Since the proposed vPresent is a cloud based system, the customers are able to access the customized virtual products from anywhere at any time, via desktop, laptop, or even smart phone. The proposed vPresent is expected to effectively deliver 3D visual information to customers and provide an interactive design platform for the development of customized products.

Free viewpoint video (FVV) provides immersive experiences in a truly seamless environment. Cloud computing facilitates the possibility to watch the FVV on mobile devices through the remote rendering in which the synthesis view is rendered in cloud server and transmitted to mobile devices. However, how to reduce the interaction delay during the viewpoint switching is a challenging problem in the remote rendering. In this paper, we propose a low-delay cloud-based FVV rendering framework to support FVV on mobile devices with satisfactory video quality and low interaction delay. In our framework, the rendering allocation scheme is proposed in which the local rendering is introduced on mobile devices during the viewpoint switching to conceal the interaction delay. To support the local rendering, the side information is generated based on the viewpoint prediction and 3D warping rule in cloud and then compressed by standard video codec. The experiment results show that the proposed remote rendering framework can improve the quality of experience substantially with improved free viewpoint video quality and low interaction delay on mobile devices.

Panorama view provides people an informative and natural user experience to represent the whole scene. The advances on mobile augmented reality, mobile-cloud computing, and mobile internet can enable panorama view on mobile phone with new functionalities, such as anytime anywhere query where a landmark picture is and what the whole scene looks like. To generate and explore panorama view on mobile devices faces significant challenges due to the limitations of computing capacity, battery life, and memory size of mobile phones, as well as the bandwidth of mobile Internet connection. To address the challenges, this paper presents a novel cloud-based mobile panorama view system that can generate and view panorama-view on mobile devices from a single picture, namely mPano". In our system, first, we propose a novel iterative multi-modal image retrieval (IMIR) approach to get spatially adjacent images using both tag and content information from the single picture. Second, we propose a cloud-based parallel server synthing approach to generate panorama view in cloud, against today's local-client synthing approach that is almost impossible for mobile phones. Third, we propose predictive-cache solution to reduce latency of image delivery from cloud server to the mobile client. We have built a real mobile panorama view system and perform experiments. The experimental results demonstrated the effectiveness of our system and the proposed key component technologies, especially for landmark images.

We present a system and a method to segment the head-shoulder image of participants in a video call using mobile devices such as a smartphone or a tablet. Participants can choose to send only the segmented head-shoulder foreground image and overlay it on top of a static background image or a background video on the receiver side of the video call, as well as to replace the background of the caller himself/herself during the video call. Our proposed method extracts the head-shoulder area of each video frame based on detected face region, superpixel clustering, and efficient label propagation.

In this paper, we propose an eye tracking and gaze estimation system for mobile phone. We integrate an eye detector, cornereye center and iso-center to improve pupil detection. The optical flow information is used for eye tracking. We develop a robust eye tracking system that integrates eye detection and optical-flow based image tracking. In addition, we further incorporate the orientation sensor information from the mobile phone to improve the eye tracking for accurate gaze estimation. We demonstrate the accuracy of the proposed eye tracking and gaze estimation system through experiments on some public video sequences as well as videos acquired directly from mobile phone.

Video consumption patterns continue to change with consumers relying more and more on on-demand Internet video and portable devices rather than traditional TV services. This new form of video service delivery and consumption makes possible more interactive and social experiences for video consumers, commonly referred to as Social TV services. This paper presents an overview of technologies and guidelines for the development of Social TV applications. A prototype using three core technologies, WebRTC, DASH, and WebSocket was developed to understand the challenges and demonstrate the feasibility of such applications.

Panel sharpening here refers to the sharpening of RGB content before displaying the content on panel, which has been frequently requested by different smartphone/tablet vendors. Although sharpening is a long existing topic, a few aspects of panel sharpening remain unclear, e.g., the optimal color space to perform sharpening for RGB input. In this study, we use unsharp mask to analyze the best color space for sharpening in terms of artifacts, light response, and multiple-to-one mapping probability. In short, the purpose of this study is to provide both theoretical analysis and real use cases for practical implementation.

The ultimate goal of network resource allocation for video teleconferencing is to optimize the Quality of Experience (QoE) of the video. The IPPP video coding structure with macroblock intra refresh is widely used for video teleconferencing. With such video coding structure, the loss of a frame generally causes error propagation to subsequent frames. A resource allocation decision of a communication network determines the QoE given that other conditions such as viewing conditions are fixed. Therefore, to optimize the QoE, a communication network needs to be able to accurately predict the QoE for each of its resource allocation decisions and then selects the decision corresponding to the best QoE. In our previous work, we reduced the QoE prediction problem to one of predicting the per-frame PSNR time series. The accuracy of the proposed per-frame PSNR prediction method was demonstrated, however, only for low resolution video sequences. In this paper, we show via simulations that the per-frame PSNR prediction method achieves good performance for higher resolution video sequences as well.

We describe an implementation of DASH streaming client for mobile devices which uses adaptation to user behavior and viewing conditions as means for improving efficiency of streaming delivery. Proposed design relies on sensors in a mobile device to detect presence of the user, his proximity to the screen, and other factors such as motion, brightness of the screen and ambient lighting conditions. This information is subsequently used to select stream that delivers adequate resolution implied by viewing conditions and natural limits of human vision. We show that in a mobile environment such adaptation can result in significant reduction of bandwidth usage compared to traditional streaming systems.

The current Internet success has already changed our social and economic world and is still continuing to revolutionize the information exchange. The exponential increase of amount and types of data that is currently exchanged on the Internet represents significant challenge for the design of future architectures and solutions. This paper reviews the current status and trends in the design of solutions and research activities in the future Internet from point of view of managing the growth of bandwidth requirements and complexity of the multimedia that is being created and shared. Outlines the challenges that are present before the video coding and approaches to the design of standardized media formats and protocols while considering the expected convergence of multimedia formats and exchange interfaces. The rapid growth of connected mobile devices adds to the current and the future challenges in combination with the expected, in near future, arrival of multitude of connected devices. The new Internet technologies connecting the Internet of Things with wireless visual sensor networks and 3D virtual worlds requires conceptually new approaches of media content handling from acquisition to presentation in the 3D Media Internet. Accounting for the entire transmission system properties and enabling adaptation in real-time to context and content throughout the media proceeding path will be paramount in enabling the new media architectures as well as the new applications and services. The common video coding formats will need to be conceptually redesigned to allow for the implementation of the necessary 3D Media Internet features.

We develop an algorithm for 4-D (RGB+Depth) video segmentation targeting immersive teleconferencing ap- plications on emerging mobile devices. Our algorithm extracts users from their environments and places them onto virtual backgrounds similar to green-screening. The virtual backgrounds increase immersion and interac- tivity, relieving the users of the system from distractions caused by disparate environments. Commodity depth sensors, while providing useful information for segmentation, result in noisy depth maps with a large number of missing depth values. By combining depth and RGB information, our work signi¯cantly improves the other- wise very coarse segmentation. Further imposing temporal coherence yields compositions where the foregrounds seamlessly blend with the virtual backgrounds with minimal °icker and other artifacts. We achieve said improve- ments by correcting the missing information in depth maps before fast RGB-based segmentation, which operates in conjunction with temporal coherence. Simulation results indicate the e±cacy of the proposed system in video conferencing scenarios.

In adaptive HTTP streaming model, the HTTP server stores multiple representations of media content, encoded at different rates. It is the function of a streaming client to select and retrieve segments of appropriate representations to enable continuous media playback under varying network conditions. In this paper we describe design of a control mechanism enabling such a selection and retrieval of media data during streaming session. We also describe the architecture of a streaming client for adaptive HTTP streaming and provide simulation data illustrating the effectiveness of the proposed control mechanism for handling bandwidth fluctuations typical for TCP traffic.

This research addresses the document vs. non-document image classification problem. The ability to select images containing text from an OCR processing stream that also includes images of scenes, people, faces, etc., will eliminate unnecessary computation and free up valuable computer resources for other tasks. This is particularly true for high volume OCR systems. Fisher vectors represent images as gradients of a global generative Gaussian Mixture Model (GMM) of low level image descriptors, and exhibit state-of-the-art performance for object categorization. Gaussian supervectors represent images by soft clustering low level image descriptors according to posterior GMM mixture probabilities, optionally using MAP adaptation, and have demonstrated state-of-the-art performance for scene categorization. We compare results obtained by applying linear SVMs to Fisher vector and Gaussian supervector representations to categorize images as having only text, no text, or a mixture of text and non-text. We also report the performance of GMM-based soft versions of vectors of locally aggregated descriptors (VLAD) and Bag of Visual words (BOV).

Automatic X-ray inspection of industrial parts usually uses reference-based methods, in which a set of model images or statistics extracted from the model image set are selected as the benchmark. Based on these methods, many systems are developed and are used extensively for anomaly detection. However, the performance of these systems relies heavily on the model image set. Thus, the selection of the model images is very important. This paper presents an approach for automatically selecting a set of model images to be used in a reference-based assisted defect recognition (ADR) system for anomaly detection of turbine blades of jet engines. The proposed approach to generating a model image set is based on feature extraction. Features are extracted from callout images of ADR, including potential defect indication type, size and location. Experimental results show that the proposed approach is fast and a low false alarm rate with acceptable detection rate is ensured. Moreover, the approach is applicable to different blade types and varied views of the blade. Further validation shows that the approach can be applied to the update of the model image set, when more images are generated from new blades and the model becomes inaccurate for anomaly detection in the new images.

The research done at the Fraunhofer IOSB in Karlsruhe within the AMFIS project is focusing on a mobile system to support rescue forces in accidents or disasters. The system consists of a ground control station which has the capability to communicate with a large number of heterogeneous sensors and sensor carriers and provides several open interfaces to allow easy integration of additional sensors into the system. Within this research we focus mainly on UAV such as VTOL (Vertical takeoff and Landing) systems because of their ease of use and their high maneuverability. To increase the positioning capability of the UAV, different onboard processing chains of image exploitation for real time detection of patterns on the ground and the interfacing technology for controlling the UAV from the payload during flight were examined. The earlier proposed static ground pattern was extended by an adaptive component which admits an additional visual communication channel to the aircraft. For this purpose different components were conceived to transfer additive information using changeable patterns on the ground. The adaptive ground pattern and their application suitability had to be tested under external influence. Beside the adaptive ground pattern, the onboard process chains and the adaptations to the demands of changing patterns are introduced in this paper. The tracking of the guiding points, the UAV navigation and the conversion of the guiding point positions from the images to real world co-ordinates in video sequences, as well as use limits and the possibilities of an adaptable pattern are examined.

High Efficiency Video Coding is the latest in the series of video coding standards developed either by MPEG, or VCEG, or jointly through a collaboration of the two committees. The first version of HEVC was completed in January 2013, but was developed without specific requirements for the compression of interlace video content. Rather, the requirements for the initial version of HEVC targeted the reduction, by 50%, of the bitrate required to delivery progressive video signals at the same, or nearly the same, visual quality as achieved by current state-of-the-art video codecs. Despite the lack of formal requirements for the support of interlace scanned content, this first version of HEVC nevertheless supports interlace video formats but achieves this support in a nominal manner, without the use of specific coding tools. Interlace formats, however, continue to be the primary format used by broadcasters for the capture and delivery of video being most recently used exclusively to capture and broadcast the 2012 Summer Olympics for the entire world. This paper explores the continued importance and relevance of interlace formats for next generation video coding standards, including HEVC. The in-progress experiments and results of a formal study of HEVC for the coding of interlace content are presented.

Archival of audio-visual databases has become an important discipline in multimedia. Various defects are typ- ically present in such archives. Among those, one can mention recording related defects such as interference between audio and video signals, optical related artifacts, recording and play out artifacts such as horizontal lines, and dropouts, as well as those due to digitization such as diagonal lines. An automatic or semi-automatic detection to identify such defects is useful, especially for large databases. In this paper, we propose two auto- matic algorithms for detection of horizontal and diagonal lines, as well as dropouts that are among the most typical artifacts encountered. We then evaluate the performance of these algorithms by making use of ground truth scores obtained by human subjects.

Land-based infrared (IR) video cameras are used to detect heat plumes expelled at the initial stages of a forest fire around the clock. Two promising methods which are based on characterization and identification of IR temporal heat plume signature are discussed. The first method uses the difference of video frames from two co-located cooled long-wave infrared (CLWIR) and cooled mid-wave infrared (CMWIR) cameras. Processing the sequence of CLWIR and CMWIR frame differences allow for effective extraction of the heat plume’s “flickering” features, which are characteristic to the early stages of a forest fire. The second method involves spectral analysis of CLWIR and CLWIR videos followed by design and implementation of a pertinent temporal filter to detect fire heat plume signal. Spikes in the Fourier transform spectra of CLWIR and CLWIR videos are identified and attributed to various dynamic features in the scene being monitored. A subsequent digital bandpass filter is implemented to detect the presence of fire heat plume by isolating and identifying its main spectral components. Further experimentation is needed to optimize performance and assess sensitivity to observation height, viewing angle, range, climate, and day/night time operation.

Extensive adoption of video surveillance, affecting many aspects of our daily lives, alarms the public about the increasing invasion into personal privacy. To address these concerns, many tools have been proposed for protection of personal privacy in image and video. However, little is understood regarding the effectiveness of such tools and especially their impact on the underlying surveillance tasks, leading to a tradeoff between the preservation of privacy offered by these tools and the intelligibility of activities under video surveillance. In this paper, we investigate this privacy-intelligibility tradeoff objectively by proposing an objective framework for evaluation of privacy filters. We apply the proposed framework on a use case where privacy of people is protected by obscuring faces, assuming an automated video surveillance system. We used several popular privacy protection filters, such as blurring, pixelization, and masking and applied them with varying strengths to people's faces from different public datasets of video surveillance footage. Accuracy of face detection algorithm was used as a measure of intelligibility (a face should be detected to perform a surveillance task), and accuracy of face recognition algorithm as a measure of privacy (a specific person should not be identified). Under these conditions, after application of an ideal privacy protection tool, an obfuscated face would be visible as a face but would not be correctly identified by the recognition algorithm. The experiments demonstrate that, in general, an increase in strength of privacy filters under consideration leads to an increase in privacy (i.e., reduction in recognition accuracy) and to a decrease in intelligibility (i.e., reduction in detection accuracy). Masking also shows to be the most favorable filter across all tested datasets.

The video coding standards comparison is historically based on objective metrics that do not correlate well with subjective results. The subjective evaluations that are done at the end of each standard development process show significant differences between the objective and subjective metrics. This paper reviews the both approaches and outlines some of the possible difference resulting in these discrepancies. Additionally suggests an approach that will compensate for these differences and help streamline the development of video coding architectures in the future by closing the gap between objective and subjective evaluations.

Depth-based 3D formats are currently being developed as extensions to both AVC and HEVC standards. The availability of depth information facilitates the generation of intermediate views for advanced 3D applications and displays, and also enables more efficient coding of the multiview input data through view synthesis prediction techniques. This paper outlines several approaches that have been explored to realize view synthesis prediction in modern video coding standards such as AVC and HEVC. The benefits and drawbacks of various architectures are analyzed in terms of performance, complexity, and other design considerations. It is hence concluded that block-based VSP prediction for multiview video signals provides attractive coding gains with comparable complexity as traditional motion/disparity compensation.

3D-AVC being developed under Joint Collaborative Team on 3D Video Coding (JCT-3V) significantly outperforms the Multiview Video Coding plus Depth (MVC+D) which simultaneously encodes texture views and depth views with the multiview extension of H.264/AVC (MVC). However, when the 3D-AVC is configured to support multiview compatibility in which texture views are decoded without depth information, the coding performance becomes significantly degraded. The reason is that advanced coding tools incorporated into the 3D-AVC do not perform well due to the lack of a disparity vector converted from the depth information. In this paper, we propose a disparity vector derivation method utilizing only the information of texture views. Motion information of neighboring blocks is used to determine a disparity vector for a macroblock, so that the derived disparity vector is efficiently used for the coding tools in 3D-AVC. The proposed method significantly improves a coding gain of the 3D-AVC in the multiview compatible mode about 20% BD-rate saving in the coded views and 26% BD-rate saving in the synthesized views on average.

In this paper, we propose a novel depth estimation method from multiple coded apertures for 3D interaction. A flat panel display is transformed into lens-less multi-view cameras which consist of multiple coded apertures. The sensor panel behind the display captures the scene in front of the display through the imaging pattern of the modified uniformly redundant arrays (MURA) on the display panel. To estimate the depth of an object in the scene, we first generate a stack of synthetically refocused images at various distances by using the shifting and averaging approach for the captured coded images. And then, an initial depth map is obtained by applying a focus operator to a stack of the refocused images for each pixel. Finally, the depth is refined by fitting a parametric focus model to the response curves near the initial depth estimates. To demonstrate the effectiveness of the proposed algorithm, we construct an imaging system to capture the scene in front of the display. The system consists of a display screen and an x-ray detector without a scintillator layer so as to act as a visible sensor panel. Experimental results confirm that the proposed method accurately determines the depth of an object including a human hand in front of the display by capturing multiple MURA coded images, generating refocused images at different depth levels, and refining the initial depth estimates.

A mobile mapping system (MMS) is a mobile multi-sensor platform developed by the geoinformation community to support the acquisition of huge amounts of geodata in the form of georeferenced high resolution images and dense laser clouds. Since data fusion and data integration techniques are increasingly able to combine the complementary strengths of different sensor types, the external calibration of a camera to a laser scanner is a common pre-requisite on today's mobile platforms. The methods of calibration, nevertheless, are often relatively poorly documented, are almost always time-consuming, demand expert knowledge and often require a carefully constructed calibration environment. A new methodology is studied and explored to provide a high quality external calibration for a pinhole camera to a laser scanner which is automatic, easy to perform, robust and foolproof. The method presented here, uses a portable, standard ranging pole which needs to be positioned on a known ground control point. For calibration, a well studied absolute orientation problem needs to be solved. In many cases, the camera and laser sensor are calibrated in relation to the INS system. Therefore, the transformation from camera to laser contains the cumulated error of each sensor in relation to the INS. Here, the calibration of the camera is performed in relation to the laser frame using the time synchronization between the sensors for data association. In this study, the use of the inertial relative movement will be explored to collect more useful calibration data. This results in a better intersensor calibration allowing better coloring of the clouds and a more accurate depth mask for images, especially on the edges of objects in the scene.

Autostereoscopic (AS) displays spatially multiplex multiple views, providing a more immersive experience by enabling users to view the content from different angles without the need of 3D glasses. Multiple views could be captured from multiple cameras at different orientations, however this could be expensive, time consuming and not applicable to some applications. The goal of multiview synthesis in this paper is to generate multiple views from a stereo image pair and disparity map by using various video processing techniques including depth/disparity map processing, initial view interpolation, inpainting and post-processing. We specifically emphasize the need for disparity processing when there is no depth information is available that is associated with the 2D data and we propose a segmentation based disparity processing algorithm to improve disparity map. Furthermore we extend the texture based 2D inpainting algorithm to 3D and further improve the hole-filling performance of view synthesis. The benefit of each step of the proposed algorithm is demonstrated with comparison to state of the art algorithms in terms of visual quality and PSNR metric. Our system is evaluated in an end-to-end multi view synthesis framework where only stereo image pair is provided as input to the system and 8 views are outputted and displayed in 8-view Alioscopy AS display.

Conventional image capture and display devices are prone to various form of optical artifacts. These artifacts are inherent to the non-ideal behavior of the various optical elements such as sensors, displays, lens, prisms, mirrors, light sources. These optical distortions may be corrected digitally by image warping technology. Image warping may be defined as a process of dynamically resampling a regularly spaced input image to produce a nonregular spacing output image. In this paper, one-pass algorithm for digital image warping is presented. It is based on anisotropic circularly symmetric antialiasing filtering over elliptical or rectangular footprint. It is shown that developed design provides flexibility and better image quality than known two-pass methods. The proposed algorithm has been embedded into a few display processors for optical distortion corrections both at image acquisition and displaying sides.

In this paper, we propose a novel post-alignment method. The method is both simple and effective for stereo video postproduction. A low-distortion algorithm for rectifying the epipolar lines is first introduced. Unlike traditional methods, which map the epipoles to (1,0,0) ^T directly, our method conducts it in two steps: 1) mapping the epipoles to points at infinity; 2) aligning the epipolar lines with x-axis. More specifically, by taking advantage of that commonly available stereoscopic movies are nearly aligned, our method keeps one of the stereo images unchanged, and the rectification is only applied to the other image. Besides epipolar non-parallel distortions, disparity distortion is also an important issue to consider for the stereoscopic movie. We propose a new constraint for stereoscopic video alignment such that the variations of disparities is also minimized. Experimental results have demonstrated that our method obtains better visual effect than the state-of-the-art methods.

Low complexity video coding addresses different applications, and is complementary to, video coding for delivery to the end user. Delivery codecs, such as the MPEG/ITU standards, provide very high compression ratios, but require high complexity and high latency. Some applications, by contrast, need the opposite characteristics of low complexity and low latency at low compression ratios. This paper discusses the applications and requirements of low complexity coding and, after discussing the prior art, describes the standard VC-2 (SMPTE 2042) codec, which is a wavelet codec designed for low complexity and ultra-low latency. VC-2 provides a wide range of coding parameters and compression ratios, allowing it to address applications such as texture coding, lossless and high dynamic range coding. In particular this paper describes the results for the low complexity coding parameters of 2 and 3 level Haar and LeGall wavelet kernels, for image regions of 4x4 and 8x8 pixels with both luma/color difference signals and RGB. The paper indicates the quality that may be achieved at various compression ratios and also clearly shows the benefit of coding luma and color components rather than RGB.

This paper proposes a Graphics Processing Unit (GPU) implementation of the stereo correspondence matching using Phase-Only Correlation (POC). The use of high-accuracy stereo correspondence matching based on POC makes it possible to measure accurate 3D shape of an object using stereo vision, while the drawback of POCbased approach is its high computational cost. Addressing this problem, we propose a GPU implementation of POC-based correspondence matching. Through a set of experiments using a variety of GPUs, we demonstrate that the proposed implementation exhibits high-speed and high-efficiency computation of POC-based stereo correspondence matching compared with the CPU implementation. We also apply the proposed implementation to two types of 3D measurement application and demonstrate their effectiveness.

Graphics Processing Units (GPUs) are freely programmable massively parallel general purpose processing units and thus offer the opportunity to off-load heavy computations from the CPU to the GPU. One application for GPU programming is image compression, where the massively parallel nature of GPUs promises high speed benefits. This article analyzes the predicaments of data-parallel image coding on the example of two high-throughput coding algorithms. The codecs discussed here were designed to answer a call from the Video Electronics Standards Association (VESA), and require only minimal buffering at encoder and decoder side while avoiding any pixel-based feedback loops limiting the operating frequency of hardware implementations. Comparing CPU and GPU implementations of the codes show that GPU based codes are usually not considerably faster, or perform only with less than ideal rate-distortion performance. Analyzing the details of this result provides theoretical evidence that for any coding engine either parts of the entropy coding and bit-stream build-up must remain serial, or rate-distortion penalties must be paid when offloading all computations on the GPU.

The measurement efficiency of Compressive Sensing (CS) enables the computational construction of images from far fewer measurements than what is usually considered necessary by the Nyquist– Shannon sampling theorem. There is now a vast literature around CS mathematics and applications since the development of its theoretical principles about a decade ago. Applications include quantum information to optical microscopy to seismic and hyper-spectral imaging. In the application of shortwave infrared imaging, InView has developed cameras based on the CS single-pixel camera architecture. This architecture is comprised of an objective lens to image the scene onto a Texas Instruments DLP® Micromirror Device (DMD), which by using its individually controllable mirrors, modulates the image with a selected basis set. The intensity of the modulated image is then recorded by a single detector. While the design of a CS camera is straightforward conceptually, its commercial implementation requires significant development effort in optics, electronics, hardware and software, particularly if high efficiency and high-speed operation are required. In this paper, we describe the development of a high-speed CS engine as implemented in a lab-ready workstation. In this engine, configurable measurement patterns are loaded into the DMD at speeds up to 31.5 kHz. The engine supports custom reconstruction algorithms that can be quickly implemented. Our work includes optical path design, Field programmable Gate Arrays for DMD pattern generation, and circuit boards for front end data acquisition, ADC and system control, all packaged in a compact workstation.

Tracking objects in dynamic scene is an interesting area of research and it has it's applications in many areas like surveillance, missile tracking system,virtual reality and robot vision. Objects in real world exhibit complex interactions with each other. When captured in a video signal, these interactions manifest themselves as in- tertwineing motions , occlusion and pose changes. A video tracking system should track these objects in this complex interactions smoothly . This paper presents a new joint method for tracking moving objects in outdoor and indoor environment. This joint method uses recursive Expectation-Maximization (EM) incorporated with Extended Kalman Filter (EKF) to estimate, fuse and track the object simultaneously, than doing it in two dif- ferent steps. This combined approach provides more realistic solution to the problem. Thereby, outperforming the conventional method of treating it as three di erent problems. We have tested our algorithm with standard dataset and real time video sequences collected from indoor environment. We also nd that the usage of the joint method improves the accuracy and computational cost. This method successfully tracks object with occlusions, di erent orientations and intertwining motion.

A common tracking algorithm solves several problems such as detection and localization of an object of interest in the scene, invariance to different movement directions of the object, occlusion of the target, when the target exits or reenters to the scene and so on. Nowadays the frame rates are various from 30 FPS (frames per second) to 300 FPS in special devices. Tracking algorithms can be classified as follows: point trackers, kernel trackers, and silhouette trackers. In this work, we propose a kernel tracking algorithm based on a local gradient histogram matching algorithm and prediction of the target position in video sequence frames.

A real-time tracking system based on adaptive correlation filtering and state prediction is proposed. The system is able to estimate at high-rate the position of multiple targets within the observed scene by taking into account information of past and present scene-frames. The position of the targets in the current frame is estimated with the help of a bank of composite correlation filters applied to several small regions taken from the observed scene. These small regions are updated in each frame according to information from a state predictor based on the motion model of targets in a twodimensional plane. The proposed system is implemented on a graphics processing unit to take advantage of massive parallelism. Computer simulation results obtained with the proposed system are presented and discussed in terms of tracking accuracy and real-time operation efficiency.

Telemedicine that involves sharing of digital data (i.e. physiological signals, 2D/3D images and videos) should meet the same standards of traditional healthcare in terms of usability, accessibility, efficiency, effectiveness and quality of clinical processes. All these requirements can be seen as elementary components that participate to the quality of experience (QoE) in an ad hoc medical application scenario. Although, the quality of service (QoS) in telemedicine has been quite investigated, QoE is still lacking clear definition in this context. This later should not be assimilated to QoS as it refers clearly to the experience by the user while QoS focuses mostly on the system. QoE has a potential relevancy to optimize and understand the technical transmission chain from the final task point viewpoint which one of the most important factor for adoption of telemedicine. Towards this goal, QoE studies should be conducted with an appropriate methodology incorporating user involvement and digital data and their relation with QoS. As one of the first effort in this field, this paper proposes a survey of some key issues and solutions associated to QoE in telemedicine. We first present the services offered by telemedicine and underline the significance of QoE for tele-diagnosis and tele-surgery. Next, we identify and analyze the influencing factors such as application area, application purpose (emergency care, acquisition assistance, second opinion, education...), content type (data specificities depending on acquisition modality), context of use (offline/real time, interactivity...), and user's state (stress, expertise...), that have to be considered for a relevant QoE assessment in telemedicine.

Microcalcifications are tiny spots of calcium deposit that often occur in female breasts. Microcalcifications are common in healthy woman, but they often are an early sign of breast cancer. On a mammogram; the current standard of care for breast screening; calcifications appear as tiny white dots. They may occur scattered throughout the breast or grouped in clusters. Radiologists determine the suspiciousness based upon several factors, including position, frequency, grouping, evolution compared to prior studies and shape. In this paper, we study micro-CT images of biopsy samples containing microcalcifications. The scanner delivers 3D images with a voxel size of 8.66 μm, i.e. ca. 8 times the spatial resolution of a contemporary digital mammogram. We propose an automated binary classification method of the samples, based upon shape analysis of the microcalcifications. The study is performed on a set of 50 benign and 50 malign samples preserved in paraffin. The ground truth of the classification is based upon anapathological investigation of the paraffin blocks. The results show a sensitivity, i.e. the percentage of correctly classified malign samples, of up to 98% with a specificity of 40%.

Optical mapping is a well established technique for recording monophasic action potential traces on the epicardial surface of isolated hearts. This measuring technique offers a high spatial resolution but it is sensitive towards myocardial motion. Motion artifacts occur because the mapping between a certain tissue portion sending out fluorescent light and a pixel of the photo detector changes over time. So far this problem has been addressed by suppressing the motion or ratiometric imaging. We developed a different approach to compensate the motion artifacts based on image registration. We could demonstrate how an image deformation field temporally changing with the heart motion could be determined. Using these deformation field time series for image transformation motion signals could be generated for each image pixel which were then successfully applied to remove baseline shift and compensate motion artifacts potentially leading to errors within maps of the first arrival time. The investigation was based on five different rat hearts stained with Di-4-ANEPPS.

A common complication associated with hip arthoplasty is prosthesis migration, and for most cemented components a migration greater than 0.85 mm within the first six months after surgery, are an indicator for prosthesis failure. Currently, prosthesis migration is evaluated using X-ray images, which can only reliably estimate migrations larger than 5 mm. We propose an automated method for estimating prosthesis migration more accurately, using CT images and image registration techniques. We report on the results obtained using an experimental set-up, in which a metal prosthesis can be translated and rotated with respect to a cadaver femur, over distances and angles applied using a combination of positioning stages. Images are first preprocessed to reduce artefacts. Bone and prosthesis are extracted using consecutive thresholding and morphological operations. Two registrations are performed, one aligning the bones and the other aligning the prostheses. The migration is estimated as the difference between the found transformations. We use a robust, multi-resolution, stochastic optimization approach, and compare the mean squared intensity differences (MS) to mutual information (MI). 30 high-resolution helical CT scans were acquired for prosthesis translations ranging from 0.05 mm to 4 mm, and rotations ranging from 0.3° to 3° . For the translations, the mean 3D registration error was found to be 0.22 mm for MS, and 0.15 mm for MI. For the rotations, the standard deviation of the estimation error was 0.18° for MS, and 0.08° for MI. The results show that the proposed approach is feasible and that clinically acceptable accuracies can be obtained. Clinical validation studies on patient images will now be undertaken.

Considering the importance of studying the movement of certain cardiac structures such as left ventricle and myocardial wall for better medical diagnosis, we propose a method for motion estimation and image segmentation in sequential Computed Tomography images. Two main tasks are tackled. The first one consists of a method to estimate the heart's motion based on a bio-inspired image representation model. Our proposal for optical flow estimation incorporates image structure information extracted from the steered Hermite transform coefficients that is later used as local motion constraints in a differential estimation approach. The second task deals with cardiac structure segmentation in time series of cardiac images based on deformable models. The goal is to extend active shape models (ASM) of 2D objects to the problem of 3D (2D + time) cardiac CT image modeling. The segmentation is achieved by constructing a point distribution model (PDM) that encodes the spatio-temporal variability of a training set. Combination of both motion estimation and image segmentation allows isolating motion in cardiac structures of medical interest such as ventricle walls.

In recent years, the use of Magnetic Resonance Imaging (MRI) to detect different brain structures such as midbrain, white matter, gray matter, corpus callosum, and cerebellum has increased. This fact together with the evidence that midbrain is associated with Parkinson’s disease has led researchers to consider midbrain segmentation as an important issue. Nowadays, Active Shape Models (ASM) are widely used in literature for organ segmentation where the shape is an important discriminant feature. Nevertheless, this approach is based on the assumption that objects of interest are usually located on strong edges. Such a limitation may lead to a final shape far from the actual shape model. This paper proposes a novel method based on the combined use of ASM and Local Binary Patterns for segmenting midbrain. Furthermore, we analyzed several LBP methods and evaluated their performance. The joint-model considers both global and local statistics to improve final adjustments. The results showed that our proposal performs substantially better than the ASM algorithm and provides better segmentation measurements.

In an increasing number of fields in medicine, precise and fast localisation of bony targets inside the body is essential. Up to now, exact localisation in the operation room can either be done with invasive methods like X-ray imaging and electromagnetic tracking systems, with volumetric ultrasound or by fixing the target in place. In this work, we present a new technology to directly track the position of the human skull through tissue in real time using infrared lasers. To achieve this, an experimental setup has been developed to precisely target a position on a subject's skin with an 850nm laser. The primary reflection on the skin is triangulated using a high-speed camera. Additionally, the reflections as well as in-tissue scattering are recorded with an in-beam setup of a NIR sensitive high-speed and high-resolution camera. Consequently, it is possible to record the scattering patterns specific to the composition of the tissue at the target. We have recorded MRI data of two test subjects (voxel size 0.15 x 0.15 x 1mm³) and extracted the soft tissue thickness with a semi-automatic segmentation approach. The MRI data was validated using force-controlled 2D ultrasound (tracked by an optical tracking system), from which soft tissue thickness was segmented manually. Optical measurements and MRI data were registered to determine soft tissue thickness for each measured laser target and finally used to train a support vector regression machine. Using the optical setup, we succeeded in computing the soft tissue thickness on the subjects' foreheads with sub-millimetre accuracy.

The optical flow associated with a set of digital images of a moving individual is analyzed in order to extract a gait signature. For this, invariant Hu moments are obtained for image description. A Hu Moment History (HMH) is obtained from K frames to describe the gait signature of individuals in a video. The gait descriptors are subsequences of the HMH of variable width. Each subsequence is generated by means of genetic algorithms and used for classification in a neuronal network. The database for algorithm evaluation is MoBo, and the gait classification results are above 90% for the cases of slow and fast walking and 100% for the cases of walking with a ball and inclined walking. An optical processor is also implemented in order to obtain the descriptors of the human gait.

This paper proposes a method for reconstructing 3D detailed structures of internal organs such as gastric wall from endoscopic video sequences. The proposed method consists of the four major steps: Feature-point-based 3D reconstruction, 3D point cloud stitching, dense point cloud creation and Poisson surface reconstruction. Before the first step, we partition one video sequence into groups, where each group consists of two successive frames (image pairs), and each pair in each group contains one overlapping part, which is used as a stitching region. Fist, the 3D point cloud of each group is reconstructed by utilizing structure from motion (SFM). Secondly, a scheme based on SIFT features registers and stitches the obtained 3D point clouds, by estimating the transformation matrix of the overlapping part between different groups with high accuracy and efficiency. Thirdly, we select the most robust SIFT feature points as the seed points, and then obtain the dense point cloud from sparse point cloud via a depth testing method presented by Furukawa. Finally, by utilizing Poisson surface reconstruction, polygonal patches for the internal organs are obtained. Experimental results demonstrate that the proposed method achieves a high accuracy and efficiency for 3D reconstruction of gastric surface from an endoscopic video sequence.

In this paper we present novel results on the reconstruction of stereoscopic information from a single phase-shift hologram captured using a 2:2 μm pixel-pitch CMOS camera in a holographic interferometer configuration. The low pixel-pitch camera allows the digitizing of holograms with a higher spatial-frequency than what has been reported in the literature, allowing the recording of macroscopic objects closer to the camera sensor. The reconstructed information can be visualized using 3D stereo glasses. From the perceived 3D we could identify several depth cues, including the occlusion effect which has not been easy to produce from single-aperture holography. The occlusion effect is also known to be difficult to produce from stereoscopic sources.

Holoscopic imaging, also known as integral imaging, has been recently attracting the attention of the research community, as a promising glassless 3D technology due to its ability to create a more realistic depth illusion than the current stereoscopic or multiview solutions. However, in order to gradually introduce this technology into the consumer market and to efficiently deliver 3D holoscopic content to end-users, backward compatibility with legacy displays is essential. Consequently, to enable 3D holoscopic content to be delivered and presented on legacy displays, a display scalable 3D holoscopic coding approach is required. Hence, this paper presents a display scalable architecture for 3D holoscopic video coding with a three-layer approach, where each layer represents a different level of display scalability: Layer 0 - a single 2D view; Layer 1 - 3D stereo or multiview; and Layer 2 - the full 3D holoscopic content. In this context, a prediction method is proposed, which combines inter-layer prediction, aiming to exploit the existing redundancy between the multiview and the 3D holoscopic layers, with self-similarity compensated prediction (previously proposed by the authors for non-scalable 3D holoscopic video coding), aiming to exploit the spatial redundancy inherent to the 3D holoscopic enhancement layer. Experimental results show that the proposed combined prediction can improve significantly the rate-distortion performance of scalable 3D holoscopic video coding with respect to the authors’ previously proposed solutions, where only inter-layer or only self-similarity prediction is used.

Publisher’s Note: This paper, originally published on 9/26/2013, was replaced with a corrected/revised version on 7/9/2014. If you downloaded the original PDF but are unable to access the revision, please contact SPIE Digital Library Customer Service for assistance. Significant research efforts have been invested in attempting to reliably capture and visualize holograms since their inception in 1962. However, less attention has been given to the efficient digital representation of the recorded holograms, which differ considerably from digitally recorded photographs. This paper examines the properties of recorded off-axis holograms and attempts to find a suitable sparse representation for holographic data. Results show significantly improved Bjøntegaard delta PSNR of over 4.5 dB on average within a bit-rate range of 0.125 to 2 bpp when combining the direction-adaptive discrete wavelet transform with non-standard decomposition schemes for off-axis holographic recordings; up to 7.5% reduction of file size has been achieved in the lossless case.

With the capability of achieving twice the compression ratio of Advanced Video Coding (AVC) with similar reconstruction quality, High Efficiency Video Coding (HEVC) is expected to become the newleading technique of video coding. In order to reduce the storage and transmission burden of digital holograms, in this paper we propose to use HEVC for compressing the phase-shifting digital hologram sequences (PSDHS). By simulating phase-shifting digital holography (PSDH) interferometry, interference patterns between illuminated three dimensional( 3D) virtual objects and the stepwise phase changed reference wave are generated as digital holograms. The hologram sequences are obtained by the movement of the virtual objects and compressed by AVC and HEVC. The experimental results show that AVC and HEVC are efficient to compress PSDHS, with HEVC giving better performance. Good compression rate and reconstruction quality can be obtained with bitrate above 15000kbps.

An analysis and discussion on the relevance of various wavelet schemes for hologram compression and reconstruction when the rendering configuration makes it possible to exploit selective refinements to perform a viewpointbased degraded reconstruction. It is observed that Gabor wavelet bases have better time-frequency localization as compared to Fresnelet bases and hence are well suited for view- dependent compression techniques for hologram reconstruction.

This paper discusses standardization activities for extending the capabilities of the High Efficiency Video Coding (HEVC) standard – the first edition of which was completed in early 2013. These near-term extensions are focused on three areas: range extensions (such as enhanced chroma formats, monochrome video, and increased bit depth), bitstream scalability extensions for spatial and fidelity scalability, and 3D video extensions (including stereoscopic/multi-view coding, and probably also depth map coding and combinations thereof). Standardization extensions on each of these topics will be completed by mid-2014, and further work beyond that timeframe is also discussed.

3D content is gaining popularity and the production and delivery of 3D video is now an active working item among video compression experts, content providers and the CE industry. Frame compatible stereo coding was initially adopted for the first generation of 3DTV broadcasting services for its compatibility with existing 2D decoders. However, the frame compatible solution sacrifices half of the original video resolution. In 2012, the Moving Picture Experts Group (MPEG) issued the call for proposal (CfP) for solutions that improve the resolution of frame compatible stereo 3D video signal while maintaining the backward compatibility to legacy decoders. The standardization process of the multiresolution frame compatible (MFC) stereo coding was then started. In this paper, the solution – Orthogonal Muxing Frame Compatible Full Resolution (OM-FCFR) – as a response to the CfP is introduced. In addition, this paper provides some experimental results for broadcasters to guide them in selecting operating points for MFC. It is observed that for typical broadcast bitrates, more than 0.5dB PSNR improvement can be achieved by MFC over the frame compatible solution with only 15%~20% overhead.

The new High Efficiency Video Coding Standard (HEVC) was finalized in January 2013. Compared to its predecessor H.264 / MPEG4-AVC, this new international standard is able to reduce the bitrate by 50% for the same subjective video quality. This paper investigates decoder optimizations that are needed to achieve HEVC real-time software decoding on a mobile processor. It is shown that HEVC real-time decoding up to high definition video is feasible using instruction extensions of the processor while decoding 4K ultra high definition video in real-time requires additional parallel processing. For parallel processing, a picture-level parallel approach has been chosen because it is generic and does not require bitstreams with special indication.

Visual privacy protection, i.e., obfuscation of personal visual information in video surveillance is an important and increasingly popular research topic. However, while many datasets are available for testing performance of various video analytics, little to nothing exists for evaluation of visual privacy tools. Since surveillance and privacy protection have contradictory objectives, the design principles of corresponding evaluation datasets should differ too. In this paper, we outline principles that need to be considered when building a dataset for privacy evaluation. Following these principles, we present new, and the first to our knowledge, Privacy Evaluation Video Dataset (PEViD). With the dataset, we provide XML-based annotations of various privacy regions, including face, accessories, skin regions, hair, body silhouette, and other personal information, and their descriptions. Via preliminary subjective tests, we demonstrate the flexibility and suitability of the dataset for privacy evaluations. The evaluation results also show the importance of secondary privacy regions that contain non-facial personal information for privacy- intelligibility tradeoff. We believe that PEViD dataset is equally suitable for evaluations of privacy protection tools using objective metrics and subjective assessments.

Standard instrumentation for the assessment of respiration rate is large and based on invasive method, and not suitable for daily inspection. An optical, simple and non-contact measurement method to detect human respiration rate using lowend imaging equipment is discussed. This technology is based on the visible light absorption of blood, which contains many important physiological information of the cardiovascular system. The light absorption of facial area can be indirectly reflected to gray value of the corresponding area image. In this paper, we acquire the respiration rate through the video signal captured by low-end imaging equipment. Firstly, the color CCD captures the facial area below the eyes and every frame of the video can be separated into three RGB channels. The blue channel is extracted as the research object. Then, we calculate the mean gray value for each image and draw the mean gray curve along the time. Fourier transform can get the frequency spectrogram of the graph, which is filtered through the Fourier filter. The extreme point is the value of the respiratory rate. Finally, an available interface program is designed and we have some volunteers tested. The correlation coefficient between the experimental data and the data provided by a reference instrument is 0.98. The consistency of the experimental results is very well. This technology costs so low that it will be widely used in medical and daily respiration rate measurement.

Amorphous materials are metastable, more reactive than the crystalline ones, and have to be evaluated before pharmaceutical compound formulation. Amorphicity is interpreted as a spatial chaos, and patterns of molecular aggregates of dexamethasone, D, were investigated in this paper by using fractal dimension, FD. Images having three magnifications of D were taken from an optical microscope, OM, and with eight magnifications, from a scanning electron microscope, SEM, were analyzed. The average FD for pattern irregularities of OM images was 1.538, and about 1.692 for SEM images. The FDs of the two kinds of images are less sensitive of threshold level. 3D images were shown to illustrate dependence of FD of threshold and magnification level. As a result, optical image of single scale is enough to characterize the drug amorphicity. As a result, the OM image at a single scale is enough to characterize the amorphicity of D.

Diabetic retinopathy (DR) is one of the major complications of diabetes that changes the blood vessels of the retina and distorts patient vision that finally in high stages can lead to blindness. Micro-aneurysms (MAs) are one of the first pathologies associated with DR. The number and the location of MAs are very important in grading of DR. Early diagnosis of micro-aneurysms (MAs) can reduce the incidence of blindness. As MAs are tiny area of blood protruding from vessels in the retina and their size is about 25 to 100 microns, automatic detection of these tiny lesions is still challenging. MAs occurring in the macula can lead to visual loss. Also the position of a lesion such as MAs relative to the macula is a useful feature for analysis and classification of different stages of DR. Because MAs are more distinguishable in fundus fluorescin angiography (FFA) compared to color fundus images, we introduce a new method based on curvelet transform and morphological operations for MAs detection in FFA images. As vessels and MAs are the bright parts of FFA image, firstly extracted vessels by curvelet transform are removed from image. Then morphological operations are applied on resulted image for detecting MAs.

In this study, an improved method of acquiring images of tunnels using a fisheye lens is presented. First, a forward image of the inside of a tunnel is obtained using a fisheye lens camera. A portion of this image is then transformed to a rectangular image using a coordinate transformation. To verify this proposed method, the technique was applied to images of an actual tunnel under construction. Although the transformed images obtained in this test were distorted, it is possible to improve the applicability of this method by correcting such distortions.

Adduced in this work are the results of investigation aimed at analysis of coordinate distributions for azimuths and ellipticity of polarization (polarization maps) in laser images of three types of biological phaseinhomogeneous layers (PhIL), namely: rough, ground and bulk scattering layers. To characterize polarization maps for all the types of PhIL, the authors have offered to use three groups of parameters: statistical moments of the first to fourth orders, autocorrelation functions and logarithmic dependences for power spectra related to distributions of azimuths and ellipticity of polarization inherent to PhIL laser images. Ascertained are the criteria for diagnostics and classification of PhIL optical properties. The possibilities of the local wavelet-analysis of polarization-inhomogeneous laser image of human blood plasma were considered. The criteria for the differentiation of the transformation of birefringence optical-anisotropic structures of blood plasma at different scales of their geometric dimensions were determined.

The optical model of polycrystalline networks of histological sections of rectum wall is suggested. The results of investigating the interrelation between the values of statistical (statistical moments of the 1^st-4^th order) parameters are presented. They characterize the coordinate distributions of the fourth parameter of Stokes vector of Fourier transforms of laser images of rectum wall histological sections and oncological changes. The diagnostic criteria of rectum cancer are determined.

In Super-Resolution, the combination technique of frequency domain and the improved keren's method has been applied in the sub-pixel image registration. The method proved to be accurate in movement estimation within given precision, but the registration accuracy was affected by the relative parameters. Based on the traditional method, an effective method of image registration for Super-Resolution in the paper was proposed in the paper. The proposed registration method has good performance by introducing the registration evaluation parameter. The experimental results demonstrate that the proposed method is effective for different test images, which takes into account the precision of estimation results and the computation efficiency as well.

simple and new technique for detection of ‘Preferred Retinal Locus’ (PRL) in human eye is proposed in this paper. Simple MATLAB algorithms for estimating RGB pixel intensity values of retinal images were used. The technique proved non-existence of ‘S’ cones in Fovea Centralis and also proposes that rods are involved in blue color perception. Retinal images of central vision loss and normal retina were taken for image processing. Blue minimum, Red maximum and Red+Green maximum were the three methods used in detecting PRL. Comparative analyses were also performed for these methods with patient’s age and visual acuity.

A method for real-time three-dimensional (3D) imaging based on Hilbert transform is proposed. Based on the properties of Hilbert transform and De Bruijn sequence, we design an encoding technique based on color fringe patterns to realize 3-D reconstruction of the phase distribution and range images. The calculation of phase map is implemented by using two sinusoidal fringe patterns with phase shifting 0 and π / 2 each other. Two phase-shifted fringe patterns are assigned to the red and blue channel of a color pattern, respectively. The phase unwrapping is accomplished with aid of the De Bruijn sequence pattern stored in the green channel. The experiment results show that the proposed method can not only acquire 3D data in real-time and one-shot fashion, but also obtain high-resolution and high-density range image data without any error propagation.

Quantization noise is present in all the current digital imaging systems, therefore its understanding and modeling is crucial for optimization of image reconstruction techniques. Hence, this paper deals with modeling of the quantization noise. We exploit the undecimated wavelet transform (UWT) for signal representation. We assume that the quantization noise in the spatial domain can be seen as additive, white and uniformly distributed. Hence, the UWT causes the transform of noise distribution due to weighted sum of noise samples and filter coefficients. From the known quantization step we are able to estimate suitable moments of noise uniform probability density function (PDF). These moments then could be directly evaluated in the undecimated wavelet domain using the derived equations. The presented algorithm gives the a priori information about the quantization noise and can be used for the suppression of it.

The optical model of polycrystalline networks of vitreous body histological sections is suggested. The results of investigating the interrelation between the values of statistical (statistical moments of the 1^st-4^th order), correlation (correlation area, asymmetry coefficient and autocorrelation function excess) and fractal (dispersion of logarithmic dependencies of power spectra) parameters are presented. They characterize the coordinate distributions of phase shifts between the orthogonal components of the amplitude in the points of laser images of histological sections of vitreous body and temporal dynamics of optical anisotropy of this biological object. The diagnostic criteria of death coming prescription are determined.

The optical model of polycrystalline networks of histological sections of rectum wall is suggested. The results of investigating the interrelation between the values of statistical (statistical moments of the 1st-4th order) parameters are presented. They characterize the coordinate polarization distributions of Fourier transforms of laser images of blood plasma and oncological changes. The diagnostic criteria of rectum cancer are determined.

The paper deals with investigating the processes of laser radiation transformation by biological liquid crystals networks using the singular optics techniques. The results obtained showed a distinct correlation between the points of "characteristic" values of coordinate distributions of Mueller matrix (M_ik = 0 , ±1) elements and polarization singularities (S- and C-points) of laser speckle field. The technique of Muellermatrix singular diagnostics of pathological changes skin derma tissue.

The process of converting of laser radiation by optically anisotropic crystals of biological networks are singular in the sense of total (simultaneous) of mechanisms of orientation and phase (birefringence) anisotropy the formation of polarization-inhomogeneous field of scattered radiation. This work is aimed at developing a method of polarization selection mechanisms of blood plasma polycrystalline networks anisotropy. The relationship between statistics, correlation and fractal parameters of polarization-inhomogeneous images of blood plasma and by linear dichroism and linear birefringence of polycrystalline networks albumin and globulin was found. The criteria of differentiation and diagnostic images of polarization-inhomogeneous plasma samples of the control group (donor) and a group of patients with malignant changes of breast tissue was identified.

On the basis of the approach to the analysis of degree of consistency between the states of polarization between two points of coherent laser field proposed by Ellis and Dogario, new parameter describing the polarization correlation similarity between manifestation of optically anisotropic components in different points of biological tissues - the degree of mutual correlation Muller matrix of biological tissue have been received. An experimental parameter in problems of differentiation orientation-phase component of architectonic and physiological state of biological tissue based on statistical and correlation approach has been proposed.

The optical model of polycrystalline networks of myometium is suggested. The results of investigating the interrelation between the values correlation (correlation area, asymmetry coefficient and autocorrelation function excess) and fractal (dispersion of logarithmic dependencies of power spectra) parameters are presented. They characterize the distributions Mueller marixes elements of myometrium histological sections. The criteria of differentiation of death coming reasons are determined.

Many spectral imaging devices are commercially available and used to detect certain skin pathology; however an alternative cost-efficient device can provide an advanced spectral analisys of skin. Multispectral device for diagnosis of pigmented skin lesions was developed and tested. Possibilities to map skin chromophores using a modified low-cost digital video-microscope is discussed. It was adapted for an advanced skin microscopy and used for detailed spectral analysis of skin. The device comprises CMOS digital imaging sensor, four-colour LED illumination system and image acquisition optics. The main goal is to obtain a set of spectral images of the skin area of interest for further conversion into maps of the main skin chromophores.

Diverse educational experiences have shown greater acceptance of children to ideas related to science, compared with adults. That fact and showing great curiosity are factors to consider to undertake scientific outreach efforts for children, with prospects of success. Moreover now 3D digital images have become a topic that has gained importance in various areas, entertainment, film and video games mainly, but also in areas such as medical practice transcendental in disease detection This article presents a system model for 3D images for educational purposes that allows students of various grade levels, school and college, have an approach to image processing, explaining the use of filters for stereoscopic images that give brain impression of depth. The system is based on one of two hardware elements, centered on an Arduino board, and a software based on Matlab. The paper presents the design and construction of each of the elements, also presents information on the images obtained and finally how users can interact with the device.

Imaging of the skin to detect the signs of cancer becomes very common these days. A new noninvasive skin imaging method developed recently gives an informative image data about the skin tissue by collecting scattered polarized light reflected from a mole area. This method scans the polarization states by continuously rotating a linearly polarized light incident on the lesion and collecting the reflected sequence of images with a CCD camera. The two main wavelengths that were used in the study were 520nm (penetration is about 100μm) and 700nm (penetration is about 200μm). Methods developed in the past to diagnose suspicious moles were implemented to visual images. However, such methods cannot be employed directly to our images because of their different properties. In this paper, we analyze skin moles obtained from this system for the purpose of distinguishing cancerous from benign moles. First, we pre-process the mole image by de-noising, contrast enhancement and intensity-based segmentation of the mole region. Then, an automatic examination of the polarized images is carried out according to characteristics such as their symmetry, cross-image local contrasts and large-scale homogeneity.

Autofocus is of fundamental importance for a real time automatic system. In many microscopy applications, a desired automatic system should provide the best focused image with enough accuracy and the least computation time. During the last years, several metrics based on images have been proposed for the autofocus process. Although many of these techniques present good results, their main limitations reside in the high computation time. Recently, the development of graphics processing units (GPUs) has given place to new scientific applications oriented to diminish the computational effort of the central processing unit (CPU). This manuscript presents the parallel implementation of eight different autofocus algorithms using GPUs for microscopy applications. The main objective of the proposed manuscript is to demonstrate that the use of GPUs can speed up the computational time required to perform the mentioned techniques. The reduction of computation time achieved with the proposed implementation suggests that graphics processing units can effectively be used for autofocus in real time microscopy.

In a 3D profilometer by the fringe projection, the shadow will be produced inevitably, thus the fringes cannot be detected in the region of the shadow. In addition, a smooth surface or a metal surface produces the specular reflection, and then, no projection fringe can be recorded in the region of oversaturation on CCD. This paper reveals a proposed system for improved these defects and shows some preliminary improved 3D profiles by the proposed dual fringe projection. To obtain the profile of sample hided in the shadow and the oversaturation, this study used the dual-projection system by two projectors. This system adopted two different directions of fringe projection and illuminates them alternately, therefore, the shadow and the oversaturation produced in their corresponding locations. Two raw 3D profiles obtained from taking the dual-projection by the four-step phase-shift. A set of algorithms used to identify the pixels of the shadow and the oversaturation, and create an error-map. According to the error-map to compensate, two 3D profiles merged into an error-reduced 3D profile. We used the solder paste as a testing sample. After comparatively analyzing the 3D images obtained by our measurement system and by a contact stylus profilometer, the result shows that our measurement system can effectively reduce the error caused by shadows and oversaturation. Fringe projection system by using a projector is a non-contact, full field and quickly measuring system. The proposed dual-projection by dual-projectors can effectively reduce the shadow and the oversaturation errors and enhance the scope of application of the 3D contour detection, especially in the detection of precision structure parts with specular reflection.

In this paper, we propose a new scene-based non-uniformity correction algorithm based on FPGA. We focus on the real application and the non-uniformity performance of the IR imager system and we did the research of the defects of other non-uniformity correction algorithms which had already been reported nowdays and found out that most of them can only be used in the lab. The scene set for these algorithms often fits themselves in most cases. In real application, the scene is mostly towards one continuous monotonous direction and has different appearance of non-uniformity. We solve these problems by developing a new projection estimator for the registration with a criterion, and immigrate this algorithm into a FPGA_based hardware system. This system is fully engineered for some particular usage. We test the performance of our technology by the evaluation indexes, and demonstrate the actual effect of correcting the non-uniformity under a monotonous motion on the system. At the end of this paper, we make a conclusion and perspective of our research work.

The present work proposes a simple low cost sensor head design making use of FBG sensor, for the measurement of liquid level. The sensor head consists of a lever, a buoyancy tube and an FBG. The lever is used to transfer the buoyancy force due to change in liquid level to the FBG resulting in shift in Bragg wavelength. The Flexibility of this design enables to measure the liquid level in an open or closed tank. The arrangement shows that liquid level sensitivity is high and is 10.7pm/mm.