Jayashri BAGADE, Kulbir SINGH, Yogesh DANDAWATE

Generalized referenceless image quality assessment framework using texture energy measures and pattern strength features

Referenceless image quality assessment is a challenging and critical problem in today's multimedia applica- tions. Texture patterns in images are normally at high frequencies compared to lower ones. Due to the effect of distortions during acquisition, compression, and transmission, texture deviation artifacts are generated that cause a granular effect in the image. Other artifacts, such as blocking, affect high frequencies in an image, causing distorted edges. Combining the analysis of texture deviation and other artifacts helps in determining the quality of an image. The proposed approach uses variation in the energy of pixels to quantify the quality of an image. These variations are calculated using texture energy measures and pattern strength-based statistical features. In the proposed approach, machine learning-based classi ers are used to predict the quality score for an image. The performance of the proposed method is tested for all images ranging from pristine to poor quality from LIVE and TID2008 databases. For different distortions, results are shown to have good correlation if they lie between the predicted score and the differential mean opinion score. Results obtained with this approach are compared with other widely used referenceless approaches. It is observed that the proposed approach shows better performance in the quanti cation of the quality of an image.

PDF

___

[1] Sheikh H, Bovik A. Image information and visual quality. IEEE T Image Process 2006; 15: 430-444. [2] Wang Z, Bovik A. Modern Image Quality Assessment. New York NY, USA: Morgan & Claypool, 2006.
[3] Avcibas I, Sankur B, Sayood K. Static evaluation of image quality measures. J Electron Imaging 2002; 11: 206-223.
[4] Eskicioglu A, Fisher P. Image quality measures and their performances. IEEE T Commun 1995; 43: 2959-2965.
[5] Wang Z, Bovik A. A universal image quality index. IEEE Signal Proc Let 2002; 9: 81-84.
[6] Wang Z, Bovik A, Sheikh H, Simoncelli E. Image quality assessment: From error visibility to structural similarity. IEEE T Image Process 2004; 13: 600-612.
[7] Zhang X, Feng X, Wang W, Xue W. Edge strength similarity for image quality assessment. IEEE Signal Proc Let 2013; 20: 319-322.
[8] Li Q, Wang Z. Reduce-reference image quality assessment using divisive normalization based image representation. IEEE J Sel Top Signa 2009; 3: 202-211.
[9] Soundararajan R, Bovik A. RRED indices: reduced reference entropic differencing for image quality assessment. IEEE T Image Process 2012; 21: 517-526.
[10] Rehman A, Wang Z. Reduced-reference image quality assessment by structural similarity estimation. IEEE T Image Process 2012; 21: 3378-3389.
[11] Saad M, Bovik A, Charier C. A DCT statistics-based blind image quality index. IEEE Signal Proc Let 2010; 17: 583-586.
[12] Shahid M, Rossholm A, Lovstrom B, Zepernick H. No reference image and video quality assessment: a classi cation and review of recent approaches. EURASIP Journal on Image and Video Processing 2014; 1.
[13] Chandler D. Seven challenges in image quality assessment: past, present and future research. International Scholarly Research Notices - Signal Processing 2013; 1: 1-52.
[14] Sheikh H, Wang Z, Cormack L, Bovik A. Live Image Quality Assessment Database. Austin, TX, USA: University of Texas. Available online at http://live.ece.utexas.edu/research/quality.
[15] Ponomarenko N, Lukin V, Zelensky A, Egiazarian V, Carli M, Battisti F. ID2008-A database for evaluation of full-reference visual quality assessment metrics. Advances of Modern Radioelectronics 2009; 10: 30-45.
[16] Wang Z, Sheikh H, Bovik A. No-reference perceptual quality assessment of JPEG compressed images. In: IEEE International Conference on Image Processing; 22{25 September 2002; Rochester, NY, USA. New York, NY, USA: IEEE. pp. 477-480.
[17] Suthaharan S. No-reference visually signi cant blocking artifact metric for natural scene images. Signal Process 2002; 89: 1647-1652.
[18] Liu H, Klomp N, Henyderickx I. A no-reference metric for perceived ringing artifacts in images. IEEE T Circ Syst Vid 2010; 20: 529-539.
[19] Zhai G, Zhang W, Yang X, Lin W, Xu Y. No-reference noticeable blockiness estimation in images. Signal Process 2008; 23: 417-432.
[20] Sazzad Z, Kawayoke Y, Horita Y. No reference image quality assessment for JPEG2000 based on spatial features. Signal Process-Image 2008; 23: 257-268.
[21] Liang L, Wang S, Chen J, Ma S, Zhou D, Gao W. No reference perceptual image quality metric using gradient pro le for JPEG 2000. Signal Process-Image 2010; 25: 502-516.
[22] Ferzil R, Karam LJ. A no-reference objective image sharpness metric based on the notion of just noticeable blur. IEEE T Image Process 2009; 18: 718-728.
[23] Zhou L, Zhang Z. No reference image quality assessment based on noise, blurring and blocking effect. Optik 2014; 125: 5677-5680.
[24] Bagade J, Dandawate Y, Singh. K. No reference image quality assessment using block based features and arti cial neural network. In: 4th International Conference on Communication in Computer and Information Science; 9{11 December 2011; Vellore, India. pp. 128-138
[25] Bagade J, Singh K, Dandawate Y. No reference image quality assessment using block-based and frequency domain statistical features: a machine learning approach. International Journal of Communication Networks and Distributed Systems 2014; 12: 95-112.
[26] Gastaldo P, Zunino R, Heynderickx I, Vicario E. Objective quality assessment of displayed images by using neural networks. Signal Process-Image 2005; 20: 643-661.
[27] Suresh S, Venkatesh Babu R, Kim H. No-reference image quality assessment using modi ed extreme learning machine classi er. Applied Soft Comput 2009; 9: 541-552.
[28] Brandao T, Queluz M. No-reference image quality assessment based on DCT domain statistics. Signal Process 2008; 88: 822-833.
[29] Liu L, Liu B, Huang H, Bovik A. No reference image quality assessment based on spatial and spectral entropies. Signal Process-Image 2014; 29: 856-863.
[30] Sheikh H, Bovik A, Cormack L. No-reference quality assessment using natural scene statistics: JPEG2000. IEEE T Image Process 2005; 14: 1918-1927.
[31] Chen M, Bovik A. No-reference image blur assessment using multiscale gradient. In: IEEE Quality of Multimedia Experience; 29{31 July 2009; San Diego, CA, USA. New York, NY, USA: IEEE. pp. 70-74.
[32] Moorthy A, Bovik A. Blind image quality assessment: from natural scene statistics to perceptual quality. IEEE T Image Process 2011; 20: 3350-3364.
[33] Mittal A, Moorthy A, Bovik A. No reference image quality assessment in the spatial domain. IEEE T Image Process 2012; 21: 4695-4707.
[34] Mittal A, Soundararajan R, Bovik A. Making a completely blind image quality analyzer. IEEE Signal Proc Let 2013; 20: 209-213.
[35] Zhang Y, Moorthy A, Chandler D, Bovik A. C-DIIVINE: No reference image quality assessment based on local magnitude and phase statistics of natural scenes. Signal Process-Image 2014; 29: 725-747.
[36] Saad M, Bovik A, Charrier C. Blind image quality assessment: A natural scene statistics approach in the DCT domain. IEEE T Image Process 2012; 21: 3339-3351.
[37] Fang Y, Ma K, Wang Z, Lin W, Fang Z, Zhai G. No reference quality assessment of contrast distorted images based on natural scene statistics. IEEE Signal Proc Let 2015; 22: 838-842.
[38] Liu L, Dong H, Huang H, Bovik A. No reference image quality assessment in curvelet domain. Signal Process-Image 2014; 29: 494-505.
[39] Laws K. Textured Image Segmentation. PhD, University of Southern California, Los Angeles, CA, USA, 1980.
[40] Ojala T, Pietikainen M. Unsupervised texture segmentation using feature distributions. Pattern Recogn 1999; 32: 477-486.
[41] Wu J, Lin W, Shi G. Image quality assessment with degradation on spatial structure. IEEE Signal Proc Let 2014; 21: 437-440.
[42] Yegnanarayana B. Arti cial Neural Network. New Delhi, India: PHI Learning Ltd., 2009.
[43] Cristianini N, Shawe-Taylor J. Introduction to Support Vector Machine. Cambridge, UK: Cambridge University Press, 2000.