Abstract
Synthetic aperture radar (SAR) has been successfully used as a remote sensing tool. However, SAR images are contaminated by speckle noise and require specialized postprocessing procedures; e.g, tailored segmenters. The G <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</sub> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> distribution is a flexible model for SAR intensities because of its ability at describing heterogeneous clutters. Furthermore, applying information theory measures (e.g., entropy) to extract features in SAR imagery processing has achieved a prominent position. In this article, we derive both a closed-form expression for the G <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</sub> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> Shannon entropy and some of its mathematical properties. Consequently new entropy-based segmentation procedures for multidimensional SAR intensities-assuming independence or some dependence pattern-are also proposed. Finally, applications to real SAR imagery point out the proposed entropy-based segmenters can be more efficient than other well-defined methods, like the clustering by gamma mixture models.
Highlights
S YNTHETIC aperture radar (SAR) has been widely employed to capture and record information of geographic scenes in remote sensing applications
We assess the performance of eight segmenters resulting from the method discussed in Section IV—on these images: 1) independent Shannon entropy (SE)-based segmenter for Γ∗ with L known (IELΓ∗ ); 2) independent SE-based segmenter for Γ∗ with L estimated (IELEΓ∗ ); TABLE I VALUES OF PERFORMANCE MEASURES FOR THE FOULUM, MUNICH, AND FLEVOLAND IMAGES SEGMENTATION
4: The conditional probability of observation H S(i) belong to gth class is calculated by τg(θ|H S(i)) in
Summary
S YNTHETIC aperture radar (SAR) has been widely employed to capture and record information of geographic scenes in remote sensing applications. Despite the use of multilook processing imposes a kind of control in the speckle effect, this last becomes SAR imagery segmentation difficult [4] To outperform this issue, Zaart et al [5] have derived a histogram-based estimation method for thresholds in segmentation issues. We provide a new entropy-based segmentation paradigm for SAR imagery, discussing as special cases resulting segmenters when Γ and GI0 laws under independence or some dependence structure are used. This approach is formulated from the distribution of the stochastic entropy.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
