Image Defogging by Multiscale Depth Fusion and Hybrid Scattering Model

The season affecting the imaging of the hill station highly and all other reasons moreover time to time. The fog in image is significantly affecting weather issue. This paper compares the hybrid scattering model and multiscale fusion method. For the single scattering of light dominated pixels the single scattering physics model is used in the hybrid model and for the remaining pixels the multiple scattering physics model (MSPM) is used. The optical thickness is the basic parameter for this pixel identification. The fusion method is as an energy minimization based method that depends on spatial Markov model. The multiscale depth fusion method (ILMRF) embeds the fusion scheme into adaptive Markov regularization to achieve better estimation of depth map. The result of the multiscale fusion is better as compared to the hybrid methodology.