Research On Transform Domain Image Level Correlation Analysis Based Remote Sensing Images Fusion Method

Transformation domain image level correlation analysis remote sensing images fusion methods are proposed.Based on the wavelet transformation image level correlation analysis fusion method and the NSCT image level correlation analysis fusion method are included in this article.Image is decomposed in the first step of the wavelet transformation fusion. And then each level component is analyzed for obtaining the most related image level, and at last, the images based on wavelet direction contrast gradient are fusion. The conclusion is proved by the simulation experiment. The fusion result is that the image edge is clearer.Image level relations is analyzed in the first step of the NSCT for obtaining the most superior decomposition level based on the correlation coefficient, and then the most superior decomposition rank is determined, and at last the images based on the image multi-direction contrast gradient are fusion. The conclusion is proved by the simulation experiment. The fusion result is that the image color is better, the image edge is clearer. Objectively, all parameters are better.The innovation spot of this article is that the image level analysis is researched before image fusion processing. The blindness of the image level relation is reduced.The fusion result is better.The frequency range, the time domain as well as the extraction characteristic value method are used in this article based on the full-color image and the multi-spectrum image. The result suggests that the time domain method is the most scientific way to find the relativity of the images.