Optimization of detailed information based on retinex algorithm for image enhancement

Jing Chen

COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 478-482

School of Electrical Engineering &Mechano-Electric Engineering, Xuchang University, Xuchang, 461000, Henan, China

To solve the problem of poor processing of detailed information in image enhancement based on traditional Retinex algorithm, this paper proposes a kind of Retinex algorithm for image enhancement based on luminance block, which first introduces the background intensity to represent the stimulus intensity according to the Webb law in psychology, so as to segment the luminance block of image, and then use enhancement factors with different scales to enhance the segmented blocks; finally, after segmentation of the edge information of pixels, fuse the information of blocks in a way of proportion solution. Simulation results show that, the Retinex algorithm for image enhancement based on luminance block proposed by this paper has improved the information entropy than the original Retinex algorithm, and this method has a good effect in the application of image enhancement.