High-precision matching of multisource remote sensing images considering geometric constraints

The traditional correlation coefficient matching method can achieve good image matching results for multi-source satellite images under plain terrain conditions. However, satellite images under mountainous and other complex terrain conditions are prone to mismatches. This article proposes a global probability relaxation automatic matching algorithm that takes into account geometric constraints. Based on the correlation coefficient criterion, combined with least squares matching and coarse to fine matching strategies, the algorithm utilizes the global probability relaxation conditions considering geometric constraints to achieve high-precision matching of multi-source images. A DEM automatic extraction algorithm based on this method is designed. The results of the multi-source image matching and the automatic generation of DEM by using this method are analyzed to validate the effectiveness of the algorithm.