Hybrid methods for automatic landscape change detection in noisy data environment

We consider most widely used practical methods for land cover change detection based on remote data sensing. Based on these methods, approaches to constructing hybrid methods are proposed. Results of the experimental study of the proposed methods in the presence of noise of various types and intensities are discussed. Based on the results of the experiments, hybrid methods that allow one to achieve a better quality in automatic change detection when compared to the known methods are determined.