Research On SAR Image Processing

In the last two decades great improvements have been made in Synthetic Aperture Radar (SAR)technology. It uses a coherent microwave sensor that can penetrate clouds and operate day or night because it provides its own illumination. According to the characteristics of SAR imaging mechanisms, geometric distortion and a form of multiplicative noise, known as coherent speckle, generally corrupt the resulting image. These present problems for standard image processing techniques. In this thesis, techniques for SAR image processing are investigated systemically in terms of speckle reduction , edge detection and ant colony algorithm.Numerous adaptive filters to reduce speckle are of great concern in SAR image processing. A filter for suppressing speckle with wavelet is proposed. This method improved the traditional method. The result of the method is better than the traditional one. The segment filter method has been proposed also, whose result is better than filters directly.For edge detection of speckled image, a new approach via accumulate power edge Checking is proposed. This approach accumulates the eight probably direction, and makes power image. Experimental results obtained from SAR images confirm that the proposed approach can bring an improvement , as compared to the Sobel , Laplacian.The ant colony algorithm has been discussed. A new ant colony algorithm has been proposed, this algorithm searches from the four edge of the image instead of the all pixel in the image .It uses the condition of move with the target side, so it reduce the calculation of the ant colony algorithm.