No-reference Image Quality Assessment Based on Natural Scene Statistics in RGB Color Space
-
摘要: RGB色彩空间中各色彩分量间存在强相关性, 图像发生失真会改变各分量间的相关性. 基于此, 本文提出了一种新的通用无参考图像质量评价方法. 首先, 根据人类视觉对RGB色彩空间中绿色分量更为敏感的颜色感知特性, 提取了G分量MSCN系数及其4方向邻域系数的统计特征; 其次, 在分析RGB色彩空间中R、G及B分量间相关性的基础上, 分别计算RGB色彩空间各色彩分量及其纹理、相位间的互信息, 利用互信息作为统计特征来描述其各分量间的相关性; 进而, 结合上述统计特征, 分别利用SVR和SVC构建无参考图像质量评价模型和图像失真类型识别模型; 最后, 在LIVE、CSIQ 及TID2008图像质量评价数据库上进行了算法与DMOS (Different mean opinion score)的相关性、失真类型识别及计算复杂性等方面的实验. 实验结果表明, 本文方法的评价结果与人类主观评价具有高度的一致性, 在LIVE 数据库上的斯皮尔曼等级相关系数和皮尔逊线性相关系数均在0.942以上; 而且, 图像失真类型识别模型的识别准确率也高达93.59%, 明显高于当今主流无参考图像质量评价方法.Abstract: There are strong correlations between the color components in the RGB color space, and distorted images can change those correlations. Based on this, a novel general-purpose no-reference image quality assessment (NR-IQA) method is proposed. Firstly, according to the color perception characteristic that human vision is more sensitive to the green component in the RGB color space, the statistical features of MSCN coefficient and its four neighboring coefficients of the G component are extracted. Secondly, on the basis of the correlation analysis between R、G and B components in RGB color space, the mutual information between the color components in RGB color space, their textures and their phases are calculated respectively. The statistical features of mutual information are used to describe the correlation between the color components in the RGB color space. Moreover, based on the aforementioned statistical features, support vector regression (SVR) and support vector classifier (SVC) are used to construct a NR-IQA model and image distortion type recognition model, respectively. At last, in order to analyze the correlation with different mean opinion score (DMOS), classification accuracy and computational complexity, a large number of simulation experiments are carried out in the LIVE, CSIQ and TID2008 image quality evaluation databases. Simulation results show that this method is suitable for many common distortions and consistent with subjective assessment, and that the Spearman's rank ordered correlation coefficient (SROCC) and the Pearson's linear correlation coefficient (PLCC) in LIVE image quality evaluation database are more than 0.942. In addition, the recognition accuracy of the recognition model is up to 93.59% and significantly superior to all present-day distortion-generic NR-IQA methods.
-
[1] Brand ao T, Queluz M P. No-reference image quality assessment based on DCT domain statistics. Signal Processing, 2008, 88(4): 822-833 [2] Golestaneh S A, Chandler D M. No-reference quality assessment of JPEG images via a quality relevance map. IEEE Signal Processing Letters, 2014, 21(2): 155-158 [3] Sheikh H R, Bovik A C, Cormack L K. No-reference quality assessment using natural scene statistics: JPEG2000. IEEE Transactions on Image Processing, 2005, 14(11): 1918-1927 [4] Zhang J, Ong S H, Le T M. Kurtosis-based no-reference quality assessment of JPEG2000 images. Signal Processing: Image Communication, 2011, 26(1): 13-23 [5] Cheng Xiao-Gang, An Ming-Wei, Ruan Ya-Duan, Chen Qi-Mei. A modern image quality measurement method for blind image restoration. Acta Automatica Sinica, 2013, 39(4): 418-423 (成孝刚, 安明伟, 阮雅端, 陈启美. 基于变分的盲图像复原质量评价指标. 自动化学报, 2013, 39(4): 418-423) [6] Lu Ya-Nan, Xie Feng-Ying, Zhou Shi-Xin, Jiang Zhi-Guo, Meng Ru-Song. Non-reference quality assessment of dermoscopy images with defocus blur and uneven illumination distortion. Acta Automatica Sinica, 2014, 40(3): 480-488 (卢亚楠, 谢凤英, 周世新, 姜志国, 孟如松. 皮肤镜图像散焦模糊与光照不均混叠时的无参考质量评价. 自动化学报, 2014, 40(3): 480-488) [7] Serir A, Beghdadi A, Kerouh F. No-reference blur image quality measure based on multiplicative multiresolution decomposition. Journal of Visual Communication and Image Representation, 2013, 24(7): 911-925 [8] Oh T, Park J, Seshadrinathan K, Lee S, Bovik A C. No-reference sharpness assessment of camera-shaken images by analysis of spectral structure. IEEE Transactions on Image Processing, 2014, 23(12): 5428-5439 [9] Ye P, Doermann D. No-reference image quality assessment using visual codebooks. IEEE Transactions on Image Processing, 2012, 21(7): 3129-3138 [10] Mittal A, Moorthy A K, Bovik A C. No-reference image quality assessment in the spatial domain. IEEE Transactions on Image Processing, 2012, 21(12): 4695-4708 [11] Dong Hong-Ping, Liu Li-Xiong. No-reference image quality assessment in mutual information domain. Journal of Image and Graphics, 2014, 19(3): 484-492 (董宏平, 刘利雄. 互信息域中的无参考图像质量评价. 中国图象图形学报, 2014, 19(3): 484-492) [12] Liu L X, Liu B, Huang H, Bovik A C. No-reference image quality assessment based on spatial and spectral entropies. Signal Processing: Image Communication, 2014, 29(8): 856-863 [13] Xue W F, Mou X Q, Zhang L, Bovik A C, Feng X C. Blind image quality assessment using joint statistics of gradient magnitude and Laplacian features. IEEE Transactions on Image Processing, 2014, 23(11): 4850-4862 [14] Sang Q B, Wu X J, Li C F, Bovik A C. Blind image quality assessment using a reciprocal singular value curve. Signal Processing: Image Communication, 2014, 29(10): 1149-1157 [15] Saad M A, Bovik A C, Charrier C. Blind image quality assessment: a natural scene statistics approach in the DCT domain. IEEE Transactions on Image Processing, 2012, 21(8): 3339-3352 [16] Moorthy A K, Bovik A C. Blind image quality assessment: from natural scene statistics to perceptual quality. IEEE Transactions on Image Processing, 2011, 20(12): 3350-3364 [17] Zhang Y, Moorthy A K, Chandler D M, Bovik A C. C-DIIVINE: No-reference image quality assessment based on local magnitude and phase statistics of natural scenes. Signal Processing: Image Communication, 2014, 29(7): 725-747 [18] Liu L X, Dong H P, Huang H, Bovik A C. No-reference image quality assessment in curvelet domain. Signal Processing: Image Communication, 2014, 29(4): 494-505 [19] Li Y M, Po L M, Xu X Y, Feng L T. No-reference image quality assessment using statistical characterization in the shearlet domain. Signal Processing: Image Communication, 2014, 29(7): 748-759 [20] Lu F F, Zhao Q F, Yang G K. A no-reference image quality assessment approach based on steerable pyramid decomposition using natural scene statistics. Neural Computing and Applications, 2015, 26(1): 77-90 [21] Li Y M, Po L M, Xu X Y, Feng L T, Yuan F, Cheung C H, Cheung K W. No-reference image quality assessment with shearlet transform and deep neural networks. Neurocomputing, 2015, 154: 94-109 [22] Tsagarisv V, Anastassopoulos V. Multispectral image fusion for improved RGB representation based on perceptual attributes. International Journal of Remote Sensing, 2005, 26(15): 3241-3254 [23] Ponomarenko N N, Lukin V V, Zelensky A, Egiazarian K, Carli M, Battisti F. TID2008---A database for evaluation of full-reference visual quality assessment metrics. Advances of Modern Radioelectronics, 2009, 10: 30-45 [24] Mittal A, Soundararajan R, Bovik A C. Making a 'Completely Blind' image quality analyzer. IEEE Signal Processing Letters, 2012, 20(3): 209-212 [25] Ruderman D L. The statistics of natural images. Network: Computation in Neural Systems, 1994, 5(4): 517-548 [26] Kovesi P. Phase congruency detects corners and edges. In: Proceedings of the 7th International Conference on Digital Image Computing: Techniques and Applications. Sydney, Australia, 2003. 309-318 [27] Klotz J G, Kracht D, Bossert M, Schober S. Canalizing boolean functions maximize mutual information. IEEE Transactions on Information Theory, 2014, 60(4): 2139-2147 [28] Chang C C, Lin C C. LIBSVM: a library for support vector machines. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3): Article No. 27
点击查看大图
计量
- 文章访问数: 1733
- HTML全文浏览量: 96
- PDF下载量: 1703
- 被引次数: 0