Affiliation:
1. Ministry of Industry and Information Technology
2. Harbin Institute of Technology
Abstract
Multi-focus image fusion is a method to extend the depth of field to generate fully focused images. The effective detection of image focusing pixels and the optimization of image regions are the key to it. A method based on multidimensional structure and edge-guided correction (MSEGC) is proposed. The pixel-level focusing evaluation function is redesigned to preserve image details and non-texture regions. Edge-guided decision correction is used to suppress edge artifacts. With public data and semiconductor detection images for verification, the results show that compared with other methods, the objective evaluation is improved by 22–50%, providing better vision.
Funder
Key Research and Development Program of Heilongjiang
CGN-HIT Advanced Nuclear and New Energy Research Institute
National Natural Science Foundation of China
Subject
Atomic and Molecular Physics, and Optics