基于小波域双阈值分割的SAR图像预处理方法

Pretreatment Method of Synthetic Aperture Radar Images Based on Wavelet Domain Dual Threshold Value Segmentation

为了抑制合成孔径雷达(SAR)图像中的相干斑噪声和背景杂波,克服使用单一SAR图像分割方法处理图像的缺点,提出了一种利用小波变换与双阈值分割相结合的SAR图像预处理方法,采用小波多尺度滤波方法削弱背景杂波及高频相干斑噪声,并使用简单易行且计算量小的双阈值分割法来滤除杂波。实验结果表明,与其它算法的分割结果相比较,该方法滤波效果好,图像边缘轮廓清晰,计算耗时短,是一种有效的SAR图像预处理方法,具有较强的实用性。

In order to restrain the coherent speckle noise and background clutter in synthetic aper-ture radar （SAR） image,overcome the disadvantages of using SAR image segmentation method on-ly to process the images,this paper presents a SAR image pretreatment method combined wavelet conversion with dual threshold value segmentation, adopts the method of wavelet multi-scale filte-ring to weaken the background clutter and high-frequency coherent speckle noise, and uses the dual threshold segmentation method which is simple and with little calculation amount to filter the clut-ter. Experiment results show： compared with the segmentation result of other algorithms, this method has good filtering effect,the contour of image edge is clear,the calculating time is short,it is an effective SAR image pretreatment method and has favorable practicability.