This paper studied on change detection algorithm of high resolution (HR) Synthetic Aperture Radar (SAR) images based
on multi-scale texture features. Firstly, preprocessed multi-temporal Terra-SAR images were decomposed by 2-D dual
tree complex wavelet transform (DT-CWT), and multi-scale texture features were extracted from those images. Then,
log-ratio operation was utilized to get difference images, and the Bayes minimum error theory was used to extract
change information from difference images. Lastly, precision assessment was done. Meanwhile, we compared with the
result of method based on texture features extracted from gray-level cooccurrence matrix (GLCM). We had a conclusion
that, change detection algorithm based on multi-scale texture features has a great more improvement, which proves an
effective method to change detect of high spatial resolution SAR images.