A Novel Retinex Algorithm and its Application to Fog-degraded Image Enhancement

Abstract

Fog-degraded images enhancement is an important problem in the field of image enhancement and computer vision. To overcome the halo artifact, enhance the contrast and better preserve the color of original image, a novel Multi-Scale Retinex color image enhancement method is proposed in this paper. An adaptive anisotropic Gaussian filtering method and its principle are described. The orientation of the Gaussian filter long axes is determined according to the gradient orientation in the position. The procedure of algorithm is given in this paper and applied in the fog- degraded images enhancement. Finally, by comparison with histogram equalization and Multi-Scale Retinex method, the experimental results show that the proposed method can offer better performance in fog-degraded image enhancement.

Keywords

• Image enhancement
• Fog-degraded image
• Multi-Scale Retinex
• Adaptive filter.