Chinese Optics Letters, 2010, 8 (5): 474, Published Online: May. 11, 2010
Image enhancement algorithm based on NF-ICM Download: 775次
图像处理 图像增强 交叉皮质模型 Sigmoid映射函数 100.0100 Image processing 100.2980 Image enhancement 100.4996 Pattern recognition, neural networks 330.1720 Color vision
Abstract
Utilizing the intersecting cortical model (ICM) to enhance degraded images under poor illumination is presented. As the key point, the general mapping function (MF) for image enhancement is deduced firstly on the basis of the nature-firing ICM (NF-ICM), which restrains the traditional autowave effect of blurring details and contrast. Then, the sigmoid MF is especially proposed to map the input gray-level to a more proper range for visual looking, and it solves over-enhancement and artifact by the classical logarithm one. For image enhancement application, the optimized parameters, initial threshold, and stopping condition in NF-ICM are all analyzed in detail. Simulation results prove that the proposed method has more contrasted, colorful, and good-visual performance.
Shan Gao, Cheng Li, Duyan Bi. Image enhancement algorithm based on NF-ICM[J]. Chinese Optics Letters, 2010, 8(5): 474.