Dark Color Image Segmentation Method Integrated MRF & Retinex

Abstract

Abstract: Given the interference from noise, poor quality and low identification of the night images, it is difficult to segment them by traditional color image segmentation method. Therefore, a novel dark color image segmentation method integrated MRF&Retinex was proposed. The -LPG-PCA filtering model was built to filter out noise of an image. The recovery function of the denoised image was constructed, before converting RGB color space to HSV color space, and the MRF & Retinex model was built for image enhancement and color correction operation. The denoised and enhanced color image was segmented by the MRF&FCM clustering method. The experimental results show that the algorithm can effectively enhance night images, eliminate noise as well as improve the peak signal noise ratio, and can achieve desirable results in dark color image segmentation.

Key words: dark color image segmentation, MRF, retinex, -LPG-PCA, FCM clustering

CLC Number:

TP391.41

Miao Yongchun, Cheng Yan. Dark Color Image Segmentation Method Integrated MRF & Retinex[J]. Journal of System Simulation, 2015, 27(10): 2387-2394.

References

[1] Deng Y, Manjunath B.S. Unsupervised Segmentation of Color-texture Regions in Images and Video[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162-8828), 2001, 23(8): 800-810.
[2] Comaniciu D, Meer P.Mean Shift: a Robust Approach toward Feature Space Analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162- 8828), 2002, 24(5): 603-619.
[3] Allen Y Y, John Wright, Yi Ma, et al. Unsupervised Segmentation of Natural Images via Lossy Data Compression[J]. Computer Vision and Image Understanding (S1077-3142), 2008, 110(2): 212-225.
[4] Geng Y Z, Chen J, Wang L.A Novel Color Image Segmentation Algorithm based on JSEG and Normalized Cuts[C]// 2013 6th International Congress on Image and Signal Processing (CISP). USA: IEEE, 2013, 1: 550-554.
[5] Mahmood Q, Chodorowski A, Persson M.Automated MRI Brain Tissue Segmentation based on Mean Shift and Fuzzy C-means using a Priori Tissue Probability Maps[J]. IRBM (S1959-0318), 2015, 36(3): 185-196.
[6] 董卓莉, 李磊, 张德贤. 基于两段多组件图割的非监督彩色图像分割算法[J]. 自动化学报, 2014 (6): 1223-1232.
[7] 纪则轩, 潘瑜, 陈强, 等. 无监督模糊C均值聚类自然图像分割算法[J]. 中国图象图形学报, 2011 (5): 773-783.
[8] Ganesan P, Rajini V.Value based Semi-automatic Segmentation of Satellite Images using HSV Color Space, Histogram Equalization and Modified FCM Clustering Algorithm[C]// 2013 International Conference on Green Computing, Communication and Conservation of Energy (ICGCE). USA: IEEE, 2013: 77-82.
[9] 占必超, 吴一全, 纪守新. 基于平稳小波变换和Retinex的红外图像增强方法[J]. 光学学报, 2010 (10): 2788-2793.
[10] Chao W H, Lai H Y, Shih Y Y I, et al. Correction of Inhomogeneous Magnetic Resonance Images using Multiscale Retinex for Segmentation Accuracy Improvement[J]. Biomedical Signal Processing and Control (S1746-8094), 2012, 7(2): 129-140.
[11] Zhao Y, Liu Y, Wu X, et al. Retinal Vessel Segmentation: An Efficient Graph Cut Approach with Retinex and Local Phase[J]. PLoS ONE (S1932-6203), 2015, 10(4): e0122332.
[12] Besag J.Spatial Interaction and the Statistical Analysis of Lattice Systems[J]. Journal of the Royal Statistical Society (S0172-7397), 1974, 36(2): 192-236.
[13] 秦绪佳, 王慧玲, 杜轶诚, 等. HSV色彩空间的Retinex结构光图像增强算法[J]. 计算机辅助设计与图形学学报, 2013 (4): 488-493.
[14] 赵泉华, 李玉, 何晓军. 结合几何划分技术和最大期望值/最大边缘概率算法的彩色图像分割[J]. 中国图象图形学报, 2013 (10): 1270-1278.
[15] Dante M V, Francisco J G F, Alberto J R S. A Fuzzy Clustering Algorithm with Spatial Robust Estimation Constraint for Noisy Color Image Segmentation[J]. Pattern Recognition Letters (S0167-8655), 2013, 34(4): 400-413.