Research on multi-exposure image fusion algorithm based on detail enhancement

Abstract

Machine vision detection relies on high-quality image extraction features obtained by the camera, and multi-exposure image fusion technology is a crucial method to acquire images with high dynamic range and rich details. In order to better obtain high dynamic range detail information of the entire scene to extract more image features, this paper proposes a detail enhancement multi-exposure image fusion algorithm based on the exposure function and the homomorphic filtering function to determine the weight. First, image sequences with different exposures are acquired by the camera and corrected, then the exposure function and homomorphic filtering are used to determine the weight map of each image, and finally the final fusion image is obtained by the multi-scale method. Compared with the current four advanced methods, the experimental results show that the algorithm can obtain better imaging quality and more image detail information, and obtain satisfactory visual effects. It provides a solution for the subsequent machine vision to extract richer features of the image.