Shao Zhenfeng, Bai Yun, Zhou Xiran. Improved Multi-scale Retinex Image Enhancement of UnderPoor Illumination[J]. Geomatics and Information Science of Wuhan University, 2015, 40(1): 32-39.
Citation:
Shao Zhenfeng, Bai Yun, Zhou Xiran. Improved Multi-scale Retinex Image Enhancement of UnderPoor Illumination[J]. Geomatics and Information Science of Wuhan University, 2015, 40(1): 32-39.
Shao Zhenfeng, Bai Yun, Zhou Xiran. Improved Multi-scale Retinex Image Enhancement of UnderPoor Illumination[J]. Geomatics and Information Science of Wuhan University, 2015, 40(1): 32-39.
Citation:
Shao Zhenfeng, Bai Yun, Zhou Xiran. Improved Multi-scale Retinex Image Enhancement of UnderPoor Illumination[J]. Geomatics and Information Science of Wuhan University, 2015, 40(1): 32-39.
1State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079 China;2Leador Spatial Information Technology Corporation Wuhan 430223 China
Severe atmosphere，optic，and other negative effects will result in low brightness and contract problem and there makes remote sensing image into low quality. In this paper，two kinds of algorithms based on human eye feature are analyzed with their advantages and limits.A novel optimized Retinex approach is proposed. It fuses Retinex theory and image enhancement algorithm that strengthens brightness and contrast via a color space transform. Brightness and contrast are shifted with additional image edge features while holding image hue being constant. The results from image enhancement can be more comfortable for human eye features，provide significant improvement in brightness and contrast，delivers richer image information，and avoids cross color phenomenon. The experimental data resource was a low-light-level image processed to illustrate the efficiency of our method via fineness，the hue bias exponent，entropy，and several other indexes.
DownLoad: