Image De-hazing Based on Improved Dark Channel Prior and Contrast Enhancement

IMAGE PROCESSING

Quality of pictures caught by sensors of the camera are degraded by haze. The haze evacuation, called dehazing, is ordinarily done under the physical degradation model, which requires a solution for an ill-posed inverse problem. To soothe the trouble of the reverse issue, a novel method called Dark channel Prior (DCP) was as of late proposed and has received a lot of consideration. Environmental factors, such as haze and cloudiness influence the picture quality and make it unacceptable for real time applications such as environment surveillance, which require pictures with clear perceivability. By and large, reproduction of mist free picture from a solitary info picture is very testing. Proposed DCP based picture defogging technique with enhanced map to abstain from artifacts, TM’s are obtained for RGB and YCbCr colour spaces. Three transmission maps for the R, G, and B channels are used to process a mean transmission map. In the YCbCr space, Y channel is utilized to figure the map. The two transmission maps are refined by protecting edge data for developing two middle pictures, which are given with weights to obtain dehazed yield. Dehazed image thus obtained suffers from low contrast and dimness. To overcome this, we go for contrast enhancement method. The RGB image is converted to HSV space. The luminance component, ‘v’ alone is considered for enhancement. Compute the illumination component and calculate the reflectance of the image. The illumination of the image is adjusted and it is multiplied by the reflectance image to get the adjusted luminance, which can be rescaled to 0 to 255 to obtain the enhanced luminance.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.