A Modified Retinex Algorithm Based on Wavelet Transformation

Retinex method mainly consists of two steps: estimation and normalization of illumination. The illumination is estimated as a smooth version of input image using low-pass filters. Some high-frequency components of image will inevitably be lost in the filtering processing, and images will lose details and information, correspondingly. In this paper, we propose a novel Retinex method applying concepts of wavelet transform, which decrease the damage in the detail information and preserve the edges and location details of images efficiently. As the illumination is low-frequency component compared with the reflectance, the illumination is estimated just using the coarse component in the wavelet expansion of the image. The experimental results shows that, compared with traditional Retinex Algorithm, the result images using modified Retinex Algorithm have greater entropy and average gradient, as well as smaller standard variance, which indicating that the images contain more details and information, and have more uniform illumination.