Synthetic Aperture Radar Images For Target Classification

Synthetic Aperture Radar (SAR) is the production of modern science and technology. It brings a great change in the functions of radars. SAR has been widely used in the areas of military and civil, it is more useful and has a supreme future. More and more researchers are engaged in this area.This paper is just about the texture features analysis of SAR images, which is a key technology in the application of SAR images. Based on data of SAR images which have been pretreated, we apply the Gray-Level Co-Occurrence Matrix method, and particularly study some texture features used for the classification of SAR images, including Difference Variance Difference Averages Difference Entropy Contrasts Energy s Variance Sum Variances Inverse Difference Moment and Correlation etc. Furthermore we have abstracted features of SAR images.If the dimension of pattern space is too big, it would result in the difficulty on calculation , and confusing with the indetermination because of high correlation of many variables. Therefore procedure of feature selection is needed. Using criterion called distance inside classes and between classes, we can get a few texture features of a SAR image which is good at image classification.With these texture features for input, making use of B-P neural network of three layers , we can proceed imageclassification. Comparing with Bayes method-the classical algorithm, we conclude that the neural network is better than Bayes method.This paper gives all the procedures of SAR image classification. The programs are all passed with amicable interface and easy operation.