Retinex Theory-based Image Enhancement Algorithms

The Retinex is one kind of image enhancement theory and it eatablished at the scientific experiment and the scientific analysis. It essence is to 'skinning'the illumination in the original image and to yield the reflectance image, that is, the original appearance of the object. Compares with conventional enhancement methods, the Retinex has shown itself to be a very versatile automatic method and can simultaneously provide sharpening, color constancy, dynamic range compression and color rendition.This thesis first discussed several basic image enhancement methods, and analyzed the shortcomings of them. Then the global and local Retinex algorithms were investigated in detail. This thesis made two modifications to the global Retinex algorithm to get good results. It discussed the relation between the local Retinex algorithm and homomorphic filtering. Afterwards it pointed out the Gaussian filter radius should be three times of the standard deviation. During the experiment, several effective parameters for image enhancement appraisal have been analyzed.Finally, this thesis proposed a combined Retinex method integrated the global and local aspects. The experiment had proven the combined method proposed in this thesis can reach the same level with the local Retinex at contrast and the entropy aspect, but can be more bright than the local Retinex.