The most used in practice land cover change detection methods by remote sensing data are considered. The approaches to the hybrid methods development those involve different methods of combining procedures and results are proposed. The results of change detection methods in different noisy data environment and intensities are presented. It is shown that an application of the hybrid methods for the change detection by data with different characteristics and noises is one of the most promising approaches to the land cover change detection, not only increasing the robustness of the results, but also simplifying the automated solution of this problem.