Adaptive Enhancement for Low-Contrast Color Images via Histogram Modification and Saturation Adjustment

Abstract

Images taken under dim or bright illumination are prone to suffer from degenerated details because of contrast compression. In order to improve the visual quality of low-contrast color images, both global and local contrast should be manipulated accordingly. For global contrast enhancement, a traditional and effective method is histogram equalization (HE). However, HE-based algorithms are prone to be badly affected by the spikes in histogram. Besides, applying HE algorithms to color images is complicated to gain vivid colors, and color distortion often occurs when color channels are processed independently. In this paper, a histogram modification method assisted by saturation and local contrast adjustment is proposed to enhance low-contrast color images. First, in order to improve global contrast, RGB histograms are redistributed according to the relative strength of RGB channels. Then, color saturation and local contrast are promoted based on the relative strength between neighboring pixels and the color-opponent mechanism. Compared with several recently published algorithms, experimental results confirm that the proposed approach produces vivid details and has the merits of protecting image naturalness and consistent regions.