Dark Channel Prior and Global Contrast Stretching based Hybrid Defogging Image Technique

Abstract

Since the foggy images undergo from low contrast and resolution due to diffusion of light and poor visibility conditions. So, Fog elimination is extremely preferred in both estimation picture making and computer vision applications. Proposed technique uses a Dark Channel Prior with contrast stretching to remove fog and improve the contrast of fog free image, respectively. Using Dark Channel Prior method one can directly take away the thickness of the haze and recover a high quality haze free image. Contrast Stretching is applied in the resulted image of dark channel prior method to improve the contrast of image. The noise that affect foggy image can also be ease by using the median low pass filter. By using this technique the visual quality and color of the foggy image can be correct effectively. Experiments are conducted on PSNR and RMSE parameters. Experimental Result shows that proposed method contains least average RMSE values and Higher PSNR values among other methods.