KEYWORDS: Clouds, Cameras, Image registration, RGB color model, Feature extraction, Detection and tracking algorithms, Data modeling, Data conversion, 3D modeling, 3D image processing
The RGB-D camera can simultaneously acquire the color and depth information of the target surface, and has been widely used in 3D modeling, machine vision and other related fields. The traditional point cloud registration algorithm only considers the geometric information, it’s operating efficiency is low and the initial value requirement is high. This paper presents a new approach to align different frames point cloud obtained by RGB-D camera, which considers visual textures and geometric information simultaneously. Firstly, detect and match feature points on RGB images, and use RANSAC algorithm to eliminate the wrong matches. Then, convert the 2d matching pairs to 3d feature point cloud based on the depth camera model, and these point pairs without deep data are deleted. Finally, calculate the camera pose parameters by performing the iterative closest point(ICP) on feature point cloud, and apply the calculated pose parameters to the whole frame data. The experimental results show that, (1) In SIFT, SURF, and ORB feature point extraction operators, ORB has the best performance for point cloud registration. (2) The proposed algorithm has a high registration accuracy, the rotation and transform estimation error are less than 0.0097 and 4.2mm respectively. (3) The algorithm also significantly improves the convergence speed, only require 0.138 seconds, and it can meet the real-time processing requirements. (4) The algorithm is insensitive to the initial values and has strong robustness.
How the different transformation models take effects on the registration accuracy based-on implicit similarity between the remote sensing images is the key point of this paper. For registration between SAR and optical imagery, analyze the imaging characteristic of push-broom optical satellite image and SAR image according to their imaging models; study the impacts taken by terrain fluctuation and different transformation models. The DEM and image pairs are simulated in the experiment, the results show: in region of bigger relief, the larger the registration image size, the greater impacts are taken by different transformation models on registration accuracy. Considering the polynomial transformation model leads to the low searching efficiency, affine transformation model regards as the best model for registration, but it has low accuracy and just applies to small images(such as 200x200). For large image (such as 800x800), 8-parameters transformation model is the best choice (balance accuracy and efficiency), but adding the parameters of transformation model (such as 12-parameters) again cannot significantly improve the registration accuracy.
This paper contrapose the airborne stripe SAR image correction, a new method was discussed which utilized the junction
condition in overlap area of adjacent stripe SAR images to assist determination the coefficients of rectification function.
At first, the rectification function style was chosen, considering the area of every stripe SAR image is small and narrow,
the polynomial correction which has second order term of x and y is chosen. During the processing of determine
rectification function, the GCPs and homologous point are used to establish the error equations, and then the coefficients
is solved by adjustment integrated in whole area. The rectification function is utilized to every stripe image, and then the
rectified stripe images are mosaic to the orthoimage of whole interesting area after the rectification. Since the junction
condition is used, the number of GCPs can be reduced and the consistency in overlap area of adjacent stripe images is
much better. The experiment Result has verified the reliability and validity of this method, it is suitable the area with few
GCPs.
This paper analyzes the principle of liquid crystal display backlight system illuminating by light emitting diode (LED) and serial kinds of structure of light guide plate. The density of diffusion dots on the light guide plate is an important factor to determine symmetric luminance and percentage of light transmission of LGP surface. Through changing the structure of light-guide plate, we can decrease the loss of light and improve the distribution of light.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.