Non-uniform Illumination Image Enhancement Based on Retinex and Gamma Correction

Abstract

Since ambient light or reflection from the surface of the object, the phenomena of none-uniform illumination exist in images.At present, some problems may appear in the image after the non-uniform illumination image processed by conventional algorithms, such as unnatural, the compressed dynamic range of brightness and halo artifacts in strong edge.To effectively process none-uniform images, the image enhancement based on Retinex theory and gamma correction was proposed.First, the bilateral filter is used to decompose the original image into an illumination and reflectance, to avoid halo artifacts in the reflectance.Second, Gamma correction and are used to equalize the dynamic range of brightness and extended brightness.Finally, the reflectance and the illumination are re-fused to restore the detail and naturalness of the image.The experimental results show that the proposed method effectively enhances the image and maintains the naturalness of the image.The objective evaluation index and computation time are also significantly better than other algorithms.