Fast Single Image Haze Removal Method for Inhomogeneous Environment Using Variable Scattering Coefficient
MetadataShow full item record
TECH SCIENCE PRESS
Image dehazingScattering coefficientSimple color balanceInhomogeneous environment
Gupta, R., Khari, M., Gupta, V., Verdu, E., Wu, X., Herrera-Viedma, E., & Gonzalez Crespo, R. (2020). Fast single image haze removal method for inhomogeneous environment using variable scattering coefficient. Cmes-Computer Modeling in Engineering & Sciences, 123(3), 1175-1192. doi:10.32604/cmes.2020.010092
The images capture in a bad environment usually loses its fidelity and contrast. As the light rays travel towards its destination they get scattered several times due to the tiny particles of fog and pollutants in the environment, therefore the energy gets lost due to multiple scattering till it arrives its destination, and this degrades the images. So the images taken in bad weather appear in bad quality. Therefore, single image haze removal is quite a bit tough task. Significant research has been done in the haze removal algorithm but in all the techniques, the coefficient of scattering is taken as a constant according to the homogeneous atmosphere but in real time this does not happen. Therefore, this paper introduces a simple and efficient method so that the scattering coefficient becomes variable according to the inhomogeneous environment. Then, this research aims to remove the haze with the help of a fast and effective algorithm i.e., Prior Color Fading, according to the inhomogeneous environmental properties. Thereby, to filter the depth map, the authors used a weighted guided image filtering which removes the drawbacks of guided image filter. Afterwards the scattering coefficient is made variable according to the inhomogeneous atmosphere and then the Simple Color Balance Algorithm is applied so that the readability property of images can be increased. The proposed method tested on various general outdoor images and synthetic hazy images and analyzed on various parameters Mean Square Error (MSE), Root Mean Square Error (RMSE), Peak Signal to Noise Ratio (PSNR), Mean Structural Similarity (MSSIM) and the Universal Objective Quality Index (UQI). Experimental results for the proposed method show that the proposed approach provides better results as compared to the state-of-the-art haze removal algorithms.