Research on image enhancement algorithm based on Retinex theory
Zhang Xueming, Mao Jian · Proceedings of the 2020 4th International Conference on Electronic Information Technology and Computer Engineering · 2020
Retinex is a method to simulate the imaging principle of human brain visual cortex, which has been widely used in solving uneven illumination. This paper introduces the origin and development of the Retinex theory firstly. Then according to the current research status of the Retinex model is divided into three types: Path-based model, variational model and centre-surround model, and each type model is summarised. Finally, the typical application of Retinex in image enhancement is introduced, the advantages and disadvantages of the three Retinex models are outlined, and the future development direction of Retinex prospects.