增强型形态学建筑指数的建筑物变化检测

针对在高分辨率遥感影像中,利用形态学建筑指数提取建筑物时,同质性区域内部会出现噪声影响建筑物变化检测精度的问题,该文提出一种基于增强型形态学建筑指数的建筑物变化检测方法。采用增强型形态学建筑指数进行建筑物的提取,较好地去除了同质性区域内部的噪声,提高了建筑物提取精度;利用变化向量分析法获得建筑物变化检测结果,并采用LFI指数对变化检测结果进行后处理,有效地区分出建筑物对象和城市道路等地物,提高了建筑物变化检测精度。最后通过实验证明:该文算法可以有效地进行建筑物的提取和变化检测。

Building change detection based on the enhanced morphological building index

In high-resolution remote-sensing imagery,noise is existed in the homogeneous regions when using morphological building index for building extraction,and the noise will also decrease the accuracy of building change detection.To solve this problem,enhanced morphological building index was proposed in the paper.The index not only can wipe out the noise in the homogeneous regions but also can improve the detection accuracy.A change vector analysis method was used for the final building change detection,and LFI index was used to post-process the results of change detection which can effectively distinguish buildings from urban roads,thus improving the building change detection accuracy.The experimental results demonstrated the effectiveness of the proposed algorithm.