Wang Lingyun
School of Big Data Engineering, Kaili University, Guizhou, Kaili 556011, China.
Heliyon. 2024 Aug 8;10(17):e35831. doi: 10.1016/j.heliyon.2024.e35831. eCollection 2024 Sep 15.
This paper proposes an adaptive image enhancement method that aims to effectively restore the brightness, detail, and natural color of various low-illumination images. To be specific, the method first constructs the initial dual-channel illumination map of the image. Next, the optimal illumination correction coefficient is calculated by the prior information entropy of the initial illumination map, which helps to correct potentially erroneous illumination estimates. To restore the illumination, gamma correction is used with the optimal illumination correction coefficient. Finally, an improved perfect reflection constraint model is used to restore the color of the image. Both visual analysis and quantitative comparison with state-of-the-art methods demonstrate the effectiveness of the method in terms of brightness adjustment, detail recovery, and color restoration.
本文提出了一种自适应图像增强方法，旨在有效恢复各种低光照图像的亮度、细节和自然色彩。具体而言，该方法首先构建图像的初始双通道光照图。接下来，通过初始光照图的先验信息熵计算最优光照校正系数，这有助于校正潜在错误的光照估计。为了恢复光照，使用伽马校正和最优光照校正系数。最后，使用改进的完美反射约束模型来恢复图像的颜色。视觉分析和与现有方法的定量比较均证明了该方法在亮度调整、细节恢复和颜色恢复方面的有效性。
