Statistical distribution-based satellite multi-spectral image waveband registration method

The invention discloses a statistical distribution-based satellite multi-spectral image waveband registration method, which comprises the following steps: (1) performingdown-sampling on a reference image and to-be-registered images by using a gauss template function to generate not less than two layers of images with different resolutions respectively by taking any spectral image of the multi-spectral image as the reference image, and processing each to-be-registered image according to the following steps; (2) calculating the marginal probability distribution and the joint probability distribution of the reference image and the to-be-registered image on the first low-resolution image layer and calculating a negative mutual information value between the reference image and the to-be-registered image, and performing iterative optimization by taking negative mutual information as a target function to solve the minimal similarity measured value or a translation transformation parameter at the time when a specifie