KEYWORDS: Solid state lighting, Hyperspectral imaging, Visualization, Image segmentation, Image classification, Monte Carlo methods, Communication engineering, Statistical analysis, Signal to noise ratio, Machine learning
The issue of limited labeled samples is still grave in hyperspectral image (HSI) classification. Collaborative learning promotes a solution to this issue by combining active learning (AL) and semisupervised learning. However, it has been found that the performance of wrong pseudolabels added into the iteration may seriously deteriorate classification performance. To tackle this problem, we propose a reverification of pseudolabels algorithm based on superpixels segmentation, which is tripartite, including two AL selection strategies, the first-verification procedure based on three classifiers, and reverification procedure improving the correctness of pseudolabeled samples based on superpixels segmentation. Specifically, two AL strategies mainly generate training samples sets for two check classifiers, and three classifiers are main forces to implement the first-verification for unlabeled samples with predictive labels. Subsequently, a reverification procedure based on local similarity in superpixels is applied to reverify these unlabeled samples with predictive labels passing the first-verification procedure. The proposed algorithm is tested on three widely used HSI datasets and compared with three state-of-the-art collaborative learning algorithms with onefold verification. Experimental results illustrate numerically and visually the significantly superior performance of our proposed algorithm considering the spatial information to reverify the correctness of pseudolabels to unlabeled samples.
Positive-intrinsic-negative diode (PIN), avalanche photodiode (APD), and single photon avalanche diode (SPAD) are commonly used optoelectronic receivers with different characteristics in practical visible light communication system. To determine the most appropriate candidate, different signal and ambient light strength in the communication scenarios should be taken into account. APD and SPAD receivers have higher sensitivities than PIN, but there are still some defects that restrain their operation performance. APD receiver may generate excessive shot noise, whereas SPAD is limited by the dead time, afterpulsing, and smaller detection area. We initially analyze the noise characteristics of PIN, APD, and SPAD receivers. SNR models are established, and the main factors affecting the performance of APD and SPAD receivers are concluded. Then, the comparison experiments of three kinds of receivers in various transmission distances are conducted, respectively. Finally, the experimental results show that: short and mid distance conditions, PIN and APD are better choices, whereas SPAD performs better in long-distance communication. In addition, if a certain intensity of ambient light is introduced, APD is still better than SPAD even in long distance.
In order to accurately monitor the working temperature of turbine blades, a signal processing technology for simulated blades is proposed in the paper. The method includes four steps: obtain the function relationship of voltage and temperature by fitting; segment the waveform of each circle by synchronization signals; filter the noise signal by Fast Fourier Transformation (FFT) and Butterworth low-pass filter; then restructure two-dimensional temperature by mapping temperature data into the polar coordinates. Finally, compared with the true temperature measured by thermocouple, the absolute temperature error after signal processing is no more than 4oC at temperatures range from 550oC to 1000oC, providing valuable guidelines for condition monitoring and fault diagnosis.
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.