Showing 1–1 of 1 results for author: Vasconcelos, G C

A Multiple Source Hourglass Deep Network for Multi-Focus Image Fusion

Authors: Fidel Alejandro Guerrero Peña, Pedro Diamel Marrero Fernández, Tsang Ing Ren, Germano Crispim Vasconcelos, Alexandre Cunha

Abstract: Multi-Focus Image Fusion seeks to improve the quality of an acquired burst of images with different focus planes. For solving the task, an activity level measurement and a fusion rule are typically established to select and fuse the most relevant information from the sources. However, the design of this kind of method by hand is really hard and sometimes restricted to solution spaces where the opt… ▽ More Multi-Focus Image Fusion seeks to improve the quality of an acquired burst of images with different focus planes. For solving the task, an activity level measurement and a fusion rule are typically established to select and fuse the most relevant information from the sources. However, the design of this kind of method by hand is really hard and sometimes restricted to solution spaces where the optimal all-in-focus images are not contained. Then, we propose here two fast and straightforward approaches for image fusion based on deep neural networks. Our solution uses a multiple source Hourglass architecture trained in an end-to-end fashion. Models are data-driven and can be easily generalized for other kinds of fusion problems. A segmentation approach is used for recognition of the focus map, while the weighted average rule is used for fusion. We designed a training loss function for our regression-based fusion function, which allows the network to learn both the activity level measurement and the fusion rule. Experimental results show our approach has comparable results to the state-of-the-art methods with a 60X increase of computational efficiency for 520X520 resolution images. △ Less

Submitted 28 August, 2019; originally announced August 2019.