Open Science Research Excellence
%0 Journal Article
%A Harpreet Kaur and  Sudipta Majumdar
%D 2018 
%J  International Journal of Computer, Electrical, Automation, Control and Information Engineering
%B World Academy of Science, Engineering and Technology
%I International Science Index 135, 2018
%T Image Dehazing Using Dark Channel Prior and Fast Guided Filter in Daubechies Lifting Wavelet Transform Domain
%U http://waset.org/publications/10008966
%V 135
%X In this paper a method for image dehazing is proposed in lifting wavelet transform domain. Lifting Daubechies (D4) wavelet has been used to obtain the approximate image and detail images.  As the haze is contained in low frequency part, only the approximate image is used for further processing. This region is processed by dehazing algorithm based on dark channel prior (DCP). The dehazed approximate image is then recombined with the detail images using inverse lifting wavelet transform. Implementation of lifting wavelet transform has the advantage of auxiliary memory saving, fast implementation and simplicity. Also, the proposed method deals with near white scene problem, blue horizon issue and localized light sources in a way to enhance image quality and makes the algorithm robust. Simulation results present improvement in terms of visual quality, parameters such as root mean square (RMS) contrast, structural similarity index (SSIM), entropy and execution time.

%P 219 - 225