中文摘要：

提出了一种新的基于DS证据理论的SAR图像多尺度边缘检测方法。该方法利用DS证据理论融合多级尺度ROA算子输出响应,在融合过程中引入检测不确定性。并通过分析ROA算子检测门限与虚警率的关系设计基本可信度分配函数（BPAF）,使得检测不确定性在门限处达到最大值。实验通过检测结果比较以及列表分析,表明本文算法能够有效降低边缘检测不确定性。

英文摘要：

To solve the problem that usual edge detectors are inefficient when applied to SAR image because of the multiplicative speckle noise, a novel SAR image multiscale edge detection method based on DS evidence theory is proposed. This method uses DS evidence theory to fuse multiscale ROA operator responses. And detection uncertainty is introduced in fusion procession. Basic probability assignment function （BPAF） is designed based on the relationship between the probability of false alarm and detection threshold, which detection uncertainty reaches maximum value at the detection threshold. It is shown by detection results and diagrams that uncertainty of edge detection is reduced effectively.