Optimal Fusion Technique for Multi-Scale Remote Sensing Images Based on DWT and CNN

The practise of fusing multiple photographs of the same scene captured at different focal lengths into a single all-focus image is known as multifocal image fusion. Local filters are utilised in most well-known fusion algorithms to capture high-frequency data before applying various fusion rules to create fused images. By decomposing the source and fusion images into numerous states, this work uses a discrete wavelet to create high-frequency and low-frequency images. The core CNN architecture in this study includes multistate extraction features & learning in residual, resulting in a multi scale & depth pan sharpening CNN data through remote sensing. Features from the images are extracted using D W T algorithms which is pre-trained. MATLAB is used to implement the suggested DWT -based picture fusion algorithm.