中文摘要：

针对采用单一特征指标选取SAR景象匹配区不全面的缺陷,提出了一种基于多种特征属性融合的SAR景象匹配区选取方法。首先,提取边缘密度、独立像元数、信息熵、主次峰值比4个特征指标;其次,运用组合赋权法将层次分析法和熵权法融合,得到各个特征指标的组合权重;最后,根据参考图像的综合评价值,规划出最佳匹配区。实验结果表明,文中算法具有较强的鲁棒性和通用性,能够有效的选取出满足要求的SAR景象匹配区。

英文摘要：

In view of disadvantage of one-sided evaluation when using single index to select SAR scene matching area, a SAR scene matehing area selection algorithm based on multi-attribute comprehensive analysis was proposed. Firstly, four feature indexes including edge density, independent pixels, information entropy and sub max ratio were chosen and analyzed. Then, to overcome poor objectivity of traditional subjective weight and poor generality of traditional objective weight, the combination weighting method combining analytic hierarchy process（AHP） with entropy weight method was used to amend indexes＇ combination weight. Finally, the comprehensive assessment values of reference images can be obtained, and the best matching area was planned. The experimental results show that this proposed algorithm has better roughness and generality, and can select SAR scene matching areas which satisfy the requirements successfully.