The 11th Iranian and the first International Conference on Machine Vision and Image Processing

Multi-Focus Image Fusion Based on Boundary Finding Through Mathematical Morphology and Active Contour Model

Zahra Samadi, Mehdi Nooshyar, Elham Mikaeli
The 11th Iranian and the first International Conference on Machine Vision and Image Processing (MVIP 2020)

Abstract

Multi-focus image fusion is the process of unifying focus information from the partially focused images of the same scene into one single fully focused image. In this paper, a novel spatial domain-based fusion technique for multi-focus images through mathematical morphology and active contour model is proposed. In the proposed method, a local energy map on image is proposed to measure focus regions and based on it, initial decision map is calculated. Then, the boundaries between focus and defocus regions are extracted and adjusted by morphological operations and free boundary condition-based active contour model. According to the found boundaries, the source images could be denoted into fully focused or defocused regions with the same focus conditions. Then, the boundary regions are fused with fitting weights. While by selecting the regions with greater focus-measures from each pair of non-boundary regions the final focused regions can be obtained. Experimental results show that this proposed method outperforms existing state-of-the-art methods, in terms of visual and quantitative evaluations.

Keywords: Multi-focus Image Fusion, Boundary Finding, Active Contour Model, Mathematical Morphology



© 2017-2021 ISMVIP All Rights Reserved