Generalization of the Dark Channel Prior for Single Image Restoration

Images degraded by light scattering and absorption, such as hazy, sandstorm, and underwater images, often suffer color distortion and low contrast because of light traveling through turbid media. In order to enhance and restore such images, we first estimate ambient light using the depth-dependent color change. Then, via calculating the difference between the observed intensity and the ambient light, which we call the scene ambient light differential, scene transmission can be estimated. Additionally, adaptive color correction is incorporated into the image formation model (IFM) for removing color casts while restoring contrast. Experimental results on various degraded images demonstrate the new method outperforms other IFM-based methods subjectively and objectively. Our approach can be interpreted as a generalization of the common dark channel prior (DCP) approach to image restoration, and our method reduces to several DCP variants for different special cases of ambient lighting and turbid medium conditions.

I. Introduction

Images or videos captured in different conditions sometimes suffer from visibility degradation because light is scattered and absorbed with distance from the camera through turbid media, such as fog, haze, sandstorms, or water. The degradation reduces the visual quality of the images and videos and affects the performance of computer vision applications. Thus, developing an effective method to restore color and contrast for such images is desirable. Fig. 1(a)-(e) shows five different images degraded by light scattering and absorption. Fig. 1.

(a)–(e) Examples of different images degraded by light scattering and absorption. The original image (a) is from [1], (b) is from [4], (c) is from [37] (d) is from [21] and (e) is from [38].