Sar Image Feature Extraction And Weak Target Detection Method

Synthetic aperture radar (SAR) is a major achievement of the development of remote sensing technology. Because of SAR's unique advantages and great value in application, many researchers show great concern and attention for it. As the rapid development of the technology of SAR, it is easy to get a large number of high-quality high-resolution SAR images. However, as the SAR images'particular imaging mechanism and the interference of speckle, it is very difficult or even impossible to read the SAR image by people directly. So using various technological approach to interpret SAR images is particularly important. This dissertation studied the feature extraction method and weak target detection method, and proposed a new weak target detection method based on S-transform and coherent technology analysis.On the field of feature extraction method for SAR images, the dissertation analyzed three basic forms of the SAR image target in the images, and introduced weak target detection difficult problems of the SAR image. This dissertation briefly described the statistical theory, SAR image sub-aperture, non-sampling profile wave (Nonsubsampled contourlet, NSCT) transform and coherent theory of technology.On the field of weak target detection method for SAR images, this dissertation analyzed the characteristics of weak target, and introduced the weak target detection method based on constant false alarm rate(CFAR), sub-aperture coherent and Nonsubsampled Contourlet transform(NSCT transform), briefly described the three basic theory, and analyzed and made comparisons of the results of these three methods.Finally the dissertation presented a method of weak target detection method based on coherent analysis in the S-transform. It detects two weak moving target in images of the same scene by using the characteristics of the S-transform of image. This dissertation gave the simulation results of the method and analyzed the results, and the simulation results showed the effectiveness of this method. The dissertation also introduced theory of S-Transform and coherence technology. The S-transform is derived from short time Fourier transform and the wavelet transform respectively, and the one-dimensional and two-dimensional formula of S-transform were obtained.