【作者】 江兴方； 陶纯堪；

【Author】 JIANG Xing-fang1’2, TAO Chun-kan( School of Electronic Engineering & Optoelectronic Technology, NanjingUniversity of Science and Technology, Nanjing 210093, China) ( Department of Information Science, Jiangsu Polytechnic University, Changzhou 213016, China)

【机构】 南京理工大学电光学院；

【摘要】 1964年,E Land从人类视觉的亮度和颜色感知出发,提出了颜色恒常知觉的计算理论Retinex。40多年来已发展成多尺度Retinex和彩色恢复多尺度Retinex等算法,增强了图像的亮度,突出了图像中暗区域的细节信息。近年来,随着Matlab和小波分析的发展,加快了Retinex的研究进程,用Matlab实现Retinex算法,提出了小波Retinex和类小波Retinex算法等。目前,Retinex算法得到了不断地改进,正从图像彩色恢复向着连续图像增强的方向发展。更多还原

【Abstract】 In 1964, E Land conceived Retinex theory from the brightness and color perceptive of human vision. The theory was developed in the past forty years. The methods of the multi-scale Retinex (MSR), MSR with color restoration were applied to enhance the brightness and highlight in the dark area of image. With development of Matlab and wavelet analysis, the progress of Retinex algorithm was larger and larger. Retinex in Matlab, wavelet Retinex algorithm, and wavelet Retinex-like algorithm was raised. Retinex theory is making progresses in enhancement from a static image to a series of dynamic images.更多还原