基于模糊综合评判的遥感图像变化检测方法

Detection method of remote sensing image change based on fuzzy comprehensive evaluation

在此提出一种基于模糊综合评判的遥感图像变化检测方法。该方法首先对某一时相的遥感图像进行基于改进的模糊C均值聚类的图像分割;其次对图像分割结果进行区域标记和特征提取,根据特征约束条件检测到感兴趣的目标区域,同时将感兴趣的目标区域映射到另一时相的遥感图像;最后综合考虑两时相遥感图像感兴趣目标区域的光谱统计特征和纹理特征,建立模糊综合评判模型对目标区域是否发生变化做出判决。

A detection method based on fuzzy comprehensive evaluation for remote sensing image change is proposed. The segment of remote sensing image in a certain time is carried out with the modified fuzzy C-means algorithm. The region labeling and features extraction of the image segmentation result are conducted. According to the feature restriction condition, the in- teresting target area is detected and mapped into remote image of another time phace. The spectrum statistical and texture fea- tures of the interesting target area from the remote sensing image at two different time phases are considered, so as to establish a fuzzy comprehensive assessment model to detect the changes.