融合Sobel算子的SAR图像结构优化方法

SAR Image Structure Optimization Method Using Sobel Operator Fusion

针对传统滚动引导滤波算法在处理SAR图像时,面对较多琐碎纹理效果欠佳,容易产生虚假边缘的问题,提出融合Sobel算子自适应计算图像中每个像素点的梯度,反映图像的边缘强度来指导滚动引导滤波参数的调整,从而优化滚动引导滤波的结果。通过传统算法和改进算法对雷达图像处理的边缘强化和细节纹理提取的对比结果表明,提出的利用Sobel算子来进行滚动引导滤波参数的调整有着明显提升,能有效地支撑雷达图像的后续处理,为图像的强散射结构提取提供了一种新的方法。

Aiming at the problem of traditional rolling guided filtering algorithms facing many trivial texture effects and prone to false edges when processing SAR images,this paper proposed a Sobel operator to adaptively calculate the gradient of each pixel in the image to reflect the edge strength of the image to guide the adjustment of rolling guided filtering parameters and optimize the results of rolling guided filtering.The results of edge enhancement and detail texture extraction in radar image processing using traditional and improved algorithms showed that the proposed Sobel operator for adjusting rolling guided filtering parameters had a significant improvement,which could effectively support the subsequent processing of radar images and provide a new method for extracting strong scattering bands in images.