基于ROEWA和Gabor滤波的SAR图像边缘提取

A new edge detection method for SAR image based on ROEWA and gabor filter

针对合成孔径雷达(SAR)图像特有的乘性噪声和非恒虚警统计特性很难正确提取目标边缘的问题,提出了在指数加权均值比(ROEWA)算子基础上寻找自适应的最佳局域Gabor滤波器进行目标边缘提取的方法。利用Gabor滤波器具有的多方向特性确定边缘方向,然后用最大似然估计纠正错误边缘方向,重新结合视觉细胞倍频程计算出Gabor函数的最佳局域滤波参数,提取出SAR图像的正确边缘。实验表明,该方法取得很好边缘提取效果,并且后期分割出的目标更符合实际目标形态,具有较强的通用性。

Due to the presence of speckle,which can be modeled as a strong,multiplicative noise,edge detection in synthetic aperture radar（SAR） images is extremely difficult.We propose a new step-edge detector for SAR images,which is search of the best local adaptive Gabor filter based on ROEWA.Edge directions are determined by multi-direction characteristics of Gabor filter.At the same time,errors of edge directions are rectified by maximum likelihood estimation.By combination of the visual cells octave and correct edge direction,parameters of the best local Gabor filter are calculated for extract the right edge in SAR images.Experiments show that the method achieves good effect edge detection,and extraction goals are more conform to the actual goal than ever,and has a strong versatility.