Single Image Enhancement Using Gradient-Norm-Based Tone Curve for Images Captured in Wide-Dynamic-Range Scenes

Abstract

When a wide range of luminance levels exist in a scene, an image captured in the scene contains very dark and bright areas with low visibility. Various methods for improving the visibility of such images have been proposed. However, the performance of these methods is not often satisfactory owing to over-enhancement such as noise amplification. Furthermore, some methods have high computational costs. In this paper, we propose an image enhancement method for improving the visibility of images captured in wide-dynamic-range scenes. The proposed method generates a tone curve for brightness and contrast enhancements. To suppress over-enhancement, the tone curve is generated by using the gradient-norm-based weighted histogram. Applying a tone curve in a pixel-wise manner may decrease the local contrast in some regions as any tone curve has a small slope in some areas. To prevent contrast reduction, the local contrast correction for controlling the effect of the tone curve in a region-adaptive manner is introduced as the post-processing of tone curve. By applying the tone curve and local contrast correction, the proposed method can enhance the visibility of an image sufficiently while suppressing over-enhancement. In addition, each processing of the proposed method is simple and suitable for fast implementation.

Keywords

• Image enhancement
• gradient norm
• histogram
• wide-dynamic-range scenes