Retinex Image Enhancement Algorithm Based On The Theory

With the continuous optimization of digital imaging equipment and the rapid development of computer technology, digital image enhancement technologies has penetrated into many fields such as industry, military, medicine, aviation, aerospace, remote sensing, etc., which play an important role in national economic development and national security, In recent years, Retinex image enhancement algorithm which is based on the color constancy of human visual system, has proven to be a hot topic of digital image enhancement technologies. It is realized by removing the incident light in the scene through human eyes simulation and the reflection properties of the reflecting object is obtained, which can improve the contrast ratio, color constancy and color fidelity. In this paper, we conducted the following research on Retinex image enhancement algorithm:Firstly, the theory of image enhancement is studied and characteristics and limitations of conventional algorithms are analyzed.Secondly, current research on Retinex image enhancement algorithm is discussed, focusing on analysis of Retinex algorithm based on iteration and center/surround.Then, according to the independence of HSI color space and image color information, a four scale Retinex image enhancement algorithm based on HSI color space is proposed by adding the process of de-noising and global adaptive brightness adjustment, in order to enhance the luminance component. And in solving the reflected component, some light information is retained while the saturation component is adjusted.Finally, simulation results on Matlab R2010a has shown that, compared with previous multi-scale Retinex algorithms and multi-scale Retinex algorithms with color restoration, the proposed algorithm in this paper has better enhancement effect and color fidelity.