Answers are grounded in this paper and cite their sources.
For the Synthetic Aperture Radar (SAR) change detection, the traditional pixel-wise methods have poor performance in terms of regional consistence and noise resistance, especially for the greatly heterogeneous aeras (e.g., built-up areas). In this article, a change detection method based on superpixel-wise likelihood-ratio tests are presented. Firstly, assuming that SAR pixel values follow a Gamma distribution, we derive the superpixel-wise probability density function of SAR images. Secondly, we conduct the likelihood-ratio test (LRT) to test the equity of two superpixels’ parameters. Thirdly, by observing the statistical properties of superpixel aeras, we simplify the calculation of LRT. Finally, we perform binarization on the LRT results according to Wilk’s Theorem to acquire the change maps. In the experiment, this method is successfully applied to flood extraction on Sentinel-1 data, showing its superiority over those traditional pixel-wise methods.
No takes yet. Share an insight, caveat, or question.
Zheng et al. (2023) studied this question.
Synapse has enriched 7 closely related papers on similar clinical questions. Consider them for comparative context: