Research On Speckle Reduction And Edge Detection Of SAR Image

Synthetic Aperture Radar (SAR) has characteristic of all-weather, day-and-night, high-resolution and strong-transmission, so it is widely used in civilian and military fields. SAR is a coherent imaging system, so SAR images contain lots of speckle noise. In this paper, in-depth studies of the SAR images speckle noise suppression and edge detection is carried out, with the sponsors of National Natural Science Foundation of China (No. 60472072), Specialized Research Foundation for the Doctoral Program of Higher Education (No. 20040699034), Aeronautical Science Foundation of China (No. 04I53070), and Special Subject of Governing Yellow River (2004SZ01-04). The principal tasks are as follows.The paper conducted an in-depth analysis of the statistical model of speckle noise in SAR images, which was got under different imaging condition and different area. It studied the SAR image filtering methods in spatial domain and wavelet frequency domain, analyzed these filter's performance. In this paper, a speckle reduction method for SAR images based on Dual Tree Complex Wavelet Transform (DTCWT) is proposed. It first presumes the probability distribution function for complex wavelet coefficients of the original nature image, and then applies MAP theory to estimate coefficients from the noisy wavelet coefficients. Filtered image can be derived after inversed DTCWT. Due to the approximate shift-invariant and good directional selectivity of DTCWT, the proposed method can improve problems existed in Discrete Wavelet Transform (DWT), which has poor performance of transform stability and edge detection. Experimental results demonstrate that the proposed method can effectively suppress speckle noise, while preserving the edge and texture information.In this paper, the Ratio edge detection methods of SAR image are introduced in detail. And the paper presents a resident place detection method. It first use Ratio edge detection operators to get edge image, and then edge texture information analysis and seed growth method is used to get candidate region. Finally, post-processing algorithm is used to get the actual resident place. Experiments show that the algorithm achieves good results.