The invention relates to a method for detecting changes in the surrounding environment of a water body based on the forward and backward heterogeneity of remote sensing images. According to the invention, an SLIC (Simple Linear Iterative Cluster) multi-scale segmentation result of superimposed images is relatively uniform, nearly square ground feature objects can be obtained in regions such as water bodies and vegetation with uniform ground feature distribution, boundary information of ground features in the images are not damaged when regions with complicated ground feature distribution are processed, and the ground feature objects are not be too fragmented, thereby being more in line with true ground feature distribution. In addition, problems are detected through analyzing changes in the surrounding environment of the water body, and a hybrid feature space is built by organically combining spectrum, texture and index features. The method has good effects on processing of pseudo-change information, which shows that the method has good detection effects for pseudo-change information such as water quality changes and vegetation phenological changes.