Song Ki Sun, Kang Hee, Kang Moon Gi
J Opt Soc Am A Opt Image Sci Vis. 2016 Jun 1;33(6):1076-88. doi: 10.1364/JOSAA.33.001076.
In this paper, an algorithm is proposed to improve contrast and saturation without color degradation. The local histogram equalization (HE) method offers better performance than the global HE method, whereas the local HE method sometimes produces undesirable results due to the block-based processing. The proposed contrast-enhancement (CE) algorithm reflects the characteristics of the global HE method in the local HE method to avoid the artifacts, while global and local contrasts are enhanced. There are two ways to apply the proposed CE algorithm to color images. One is luminance processing methods, and the other one is each channel processing methods. However, these ways incur excessive or reduced saturation and color degradation problems. The proposed algorithm solves these problems by using channel adaptive equalization and similarity of ratios between the channels. Experimental results show that the proposed algorithm enhances contrast and saturation while preserving the hue and producing better performance than existing methods in terms of objective evaluation metrics.
本文提出了一种在不降低颜色质量的情况下提高对比度和饱和度的算法。局部直方图均衡化（HE）方法比全局HE方法具有更好的性能，而局部HE方法有时会由于基于块的处理而产生不理想的结果。所提出的对比度增强（CE）算法在局部HE方法中体现了全局HE方法的特性，以避免出现伪像，同时增强全局和局部对比度。将所提出的CE算法应用于彩色图像有两种方法。一种是亮度处理方法，另一种是各通道处理方法。然而，这些方法会导致饱和度过高或降低以及颜色质量下降的问题。所提出的算法通过使用通道自适应均衡化和通道之间的比例相似性来解决这些问题。实验结果表明，所提出的算法在保持色调的同时增强了对比度和饱和度，并且在客观评估指标方面比现有方法具有更好的性能。
