Interactive and Fast Low-Light Image Enhancement Algo-rithm and Application. Issue 1 (1st April 2022)

• Record Type: Journal Article
• Authors: Shi, Yangming; Wu, Xiaopo; Zhu, Ming
• Is Part Of: Journal of physics. Volume 2258:Issue 1(2022)
• Publication Date: 2022-04-01
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/2258/1/012003
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: To obtain personalized outcomes for the low-light image enhancement, a novel interactive algorithm based on the well-designed Gamma Curve is proposed to enrich the enhancement techniques. Different from the previous works trying to enhance the image in solely brightness or naturalness by a specific designed deep network, the proposed method is capable of controlling the output results according to the user's preferences by the same framework with different parameters. There would be three main advantages brought by the proposed network: 1) Interactivity, which allows to generate enhancements results according to users' preferences in human-interactive manners; 2) Convenience, wherein the model only needs to train for once without using any reference images, and then can obtain results with different brightness during testing by adjusting the hyper-parameter. 3) Fastness, which results from the lightweight network and the excellent properties of the Gamma Curve to make the network operate in extraordinary high speed. Experiments demonstrate the superiority of our algorithm relative to the previous work. In addition, a multi-platform low-illumination enhancement software is explored to facilitate its application for the public.