Dark channel prior based image de-hazing: A review

Digital images captured under poor environments are vulnerably degraded in their capacities to convey adequate amount of information to the viewer or computer-based processes. One of the common causes affecting the quality of outdoor images can be traced to that coming from atmospheric condensations such as fog or haze. Image processing algorithms, hence, had been developed to address the de-hazing problem in order to recover the scene information. Approaches based on the dark channel prior, in particular, had initiated a large number of research activities because of its satisfactory performance and possibilities for further improvements and applications. In this paper, a review on methods based on the dark channel prior is presented. The principle of restoration by a ray transmission model applied in image de-hazing is examined together with a classification of the models commonly employed. The difficulties encountered in the implementation of de-hazing algorithms are addressed and discussed. A summary of critical issues and a discussion of future trends are also included in this review.