Fog Removal by Multiple Polynomial Regression Model through Curvelets
Monika Verma, Vandana Dixit Kaushik, Vinay Kumar Pathak · International journal of intelligent engineering and systems · 2017
Removal of fog in a Single Image has been a challenging task.Fog causes the visibility to drop thus causing problems for the sensors to capture the scene perfectly.In this work the fog is removed by using multiple polynomial regression model through curvelets.The depth map can be obtained by using the Multiple Polynomial Regression Model and the clear image can be obtained by using the curvelets.Curvelet transform are the multiscale directional transformations which are useful in capturing the relevant information even in the presence of noise.Curvelets have a high directional sensitivity which makes them suitable for representing images even in the presence of noise.If the depth information is extracted then the scene radiance can be easily formulated by using the atmospheric scattering model.In terms of efficiency and experimental results the proposed algorithm works better than the existing approaches.