一种基于Retinex的非线性彩色图像增强算法
Nonlinear Colorful Image Enhancement Based on Retinex
DOI: 10.12677/jisp.2012.11001, PDF, HTML,    国家自然科学基金支持
作者: 余先川, 倪锋, 胡丹, 张立保, 徐金东：北京师范大学信息科学与技术学院，北京
关键词: Retinex；HVS；图像增强；非线性变换； Retinex; Hvs; Image Enhancement; Nonlinear Transform
摘要:

针对多尺度Retinex彩色恢复(MSRCR)算法容易造成颜色失真和丢失细节信息的缺点，本文提出了一种基于Retinex的非线性彩色图像增强算法。首先将图像从RGB空间转换到YCbCr空间，从而避免了直接在RGB空间进行处理时易造成颜色失真的缺陷，然后采用改进的Retinex模型进行局部自适应增强，很好地保留了图像的细节信息，再利用Gamma校正做全局亮度调整，对图像的整体动态范围进行压缩，最后再将图像从YCbCr空间转换到RGB空间。实验结果表明，算法对自然图像和遥感影像的处理结果都很有效，没有出现颜色失真现象，在改善视觉效果的同时也增强了图像的细节信息。

Abstract: To overcome the drawbacks of color distortion and losing details of multi-scale Retinex with color restoration, a nonlinear colorful image enhancement algorithm based on retinex is proposed. Firstly, the image is transformed from RGB color space to YCbCr color space, in order to avoid the defect of color distortion when it is processed directly in the RGB color space. Secondly, the image is locally adapted according to the modified Retinex model, which can retain the details well. Then, Gamma correction is used for global intensity control to compress dynamic range of the image. Finally, the result is transformed from YCbCr color space to RGB color space. Experiments show that our method has a good performance on either natural photo or remote sensing image, which reduces the color distortion, improves the visual appearance and enhances the detailed information.. Applying to remote sensing image classification, the result of the image processed in our method is more reliable and more accurate.

 

文章引用：余先川, 倪锋, 胡丹, 张立保, 徐金东. 一种基于Retinex的非线性彩色图像增强算法[J]. 图像与信号处理, 2012, 1(1): 1-7. http://dx.doi.org/10.12677/jisp.2012.11001

参考文献

[1] E. Land, J. McCann. Lightness and retinex theory. Journal of the Optical Society of America, 1971, 61(1): 1-11.
[2] E. Land. Recent advances in Retinex theory and some implications for cortical computations: Color vision and the natural image. Proceedings of the National Academy of Sciences of the United States of America, 1983, 80(16): 5163-5169.
[3] B. K. P. Horn. Determining lightness from an image. Computer Graphics and Image Processing, 1974, 3: 277-299.
[4] A. Hurlbert. Formal connections between lightness algorithms. Journal of the Optical Society of America A, 1986, 3: 1684- 1692.
[5] D. J. Jobson, Z. Rahman and G. A. Woodell. Properties and performance of a center/surround retinex. IEEE Transactions on Image Process, 1997, 6(3): 451-462.
[6] Z. Rahman, D. J. Jobson and G. A. Woodell. Image enhancement, image quality, and noise. Proceeding of SPIE, 2005, 5907, Ar- ticle ID: 59070N.
[7] Z. U. Rahman, D. J. Jobson and G. A. Woodell. Retinex pro- cessing for automatic image enhancement. Journal of Electronic Imaging, 2004, 13(1): 100-110.
[8] R. Kimmel, M. Elad, D. Shaked, et al. A variational framework for retinex. International Journal of Computer Vision, 2003, 52(1): 7-23.
[9] L. Meylan, S. Susstrunk. Color image enhancement using a Retinex-based adaptive filter. CGIV, 2004: 359-363.
[10] L. Meylan, S. Susstrunk. High dynamic range image rendering with a Retinex-based adaptive filter. IEEE Transactions on Image Processing, 2006, 15(9): 2820-2830.
[11] M. Bertalmio, J. D. Cowan. Implementing the Retinex algorithm with Wilson-Cowan equations. Journal of Physiology, Paris (France), 2009, 103(1-2): 69-72.
[12] J. M. Morel, A. B. Petro and C. Sbert. A PDE formalization of Retinex theory. IEEE Transactions on Image Process, 2010, 19(11): 2825-2837.
[13] K. Fukunaga, L. D. Hostetler. The estimation of the gradient of a density function, with application in pattern recognition. IEEE Transactions on Information Theory, 1975, 21: 32-40.
[14] Y. Cheng. Mean shift, mode seeking, and clustering. IEEE Tran- sactions on Pattern Analysis and Machine Intelligence, 1995, 17(8): 790-799.
[15] D. Comaniciu, P. Meer. Mean shift: A robust approach toward feature space analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(5): 603-619.