Weighted fusion algorithm for pixel-level multiresolution images

Abstract Weighted fusion for pixel-level multiresolution images improves the image quality. The commonly used algorithm is based on wavelet transform, but the pixel quality after fusion is poor, and the fusion process is complicated. Therefore, a weighted fusion algorithm for pixel-level multiresolution images is put forward based on chaotic neural network. Firstly, pixel-level multiresolution image fusion is analyzed, and the image is preprocessed and matched. Then through the dynamic search feature of chaotic neural network to find the global optimal solution, so as to realize the weighted fusion. Experiments show that the algorithm proposed for image fusion results in image of high resolution, and the image fusion process is simple.