Translator Disclaimer
17 December 1996 Optimum classification of correlated SAR textures
Author Affiliations +
Previous studies on discriminating between SAR clutter textures have been largely based on optimizing the performance for the single-point statistics of texture alone. This paper makes a study of the effects of introducing spatial correlations on classification into a set of predefined textures. Maximum likelihood (ML) estimation is adopted throughout. Classification based on fitting the autocorrelation function is compared with that based on the amplitude spectrum of the data. In each case, care is taken to derive a good approximation to the observed data distributions so that the ML treatment is well founded.
© (1996) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Christopher John Oliver "Optimum classification of correlated SAR textures", Proc. SPIE 2958, Microwave Sensing and Synthetic Aperture Radar, (17 December 1996);

Back to Top