A multiscale retinex based on wavelet transformation
Kozue Kawasaki, Akira Taguchi · 2014
In this paper, we propose a multi-scale retinex (MSR) in the wavelet transformation domain. Retinex method consists mainly two steps: estimation and normalization of illumination. Illumination has only low-frequency components. The estimation of illumination can be performed by using the coarse component in the wavelet expansion of input image. The output of standard MSR is a weighted sum of several different SSR outputs. Thus, the MSR algorithm has still heavy processing to deal such as motion pictures because of Gaussian filter with wide surround space is used. On the other hand, in our method, Gaussian filtering is applied to half or quarter size of the original image. Furthermore, the surround space of Gaussian filter can be set small. Thus, the computational cost of proposed method is only 1% of that of standard MSR. Moreover, the output image with higher entropy can be automatically derived by introducing a novel clipping and gain/offset operation.