摘要
Considering the relatively poor robustness of quality scores for different types of distortion and the lack of mechanism for determining distortion types, a no-reference image quality assessment(NR-IQA) method based on the Ada Boost BP neural network in the wavelet domain(WABNN) is proposed. A 36-dimensional image feature vector is constructed by extracting natural scene statistics(NSS) features and local information entropy features of the distorted image wavelet sub-band coefficients in three scales. The ABNN classifier is obtained by learning the relationship between image features and distortion types. The ABNN scorer is obtained by learning the relationship between image features and image quality scores. A series of contrast experiments are carried out in the laboratory of image and video engineering(LIVE) database and TID2013 database. Experimental results show the high accuracy of the distinguishing distortion type, the high consistency with subjective scores and the high robustness of the method for distorted images. Experiment results also show the independence of the database and the relatively high operation efficiency of this method.
Considering the relatively poor robustness of quality scores for different types of distortion and the lack of mechanism for determining distortion types, a no-reference image quality assessment(NR-IQA) method based on the Ada Boost BP neural network in the wavelet domain(WABNN) is proposed. A 36-dimensional image feature vector is constructed by extracting natural scene statistics(NSS) features and local information entropy features of the distorted image wavelet sub-band coefficients in three scales. The ABNN classifier is obtained by learning the relationship between image features and distortion types. The ABNN scorer is obtained by learning the relationship between image features and image quality scores. A series of contrast experiments are carried out in the laboratory of image and video engineering(LIVE) database and TID2013 database. Experimental results show the high accuracy of the distinguishing distortion type, the high consistency with subjective scores and the high robustness of the method for distorted images. Experiment results also show the independence of the database and the relatively high operation efficiency of this method.
引文
[1]MANAP R A,SHAO L.Non-distortion-specific no-reference image quality assessment:a survey.Information Sciences,2015,301(C):141-160.
[2]ZHU X,MILANFAR P.A no-reference sharpness metric sensitive to blur and noise.Proc.of International Workshop on Quality of Multimedia Experience,2009:64-69.
[3]CHEN M J,BOVIK A C.No-reference image blur assessment using multiscale gradient.Proc.of International Workshop on Quality of Multimedia Experience,2009:1-11.
[4]MARICHAL X,MA W Y,ZHANG H J.Blur determination in the compressed domain using DCT information.Proc.of the IEEE International Conference on Image Processing,2002,12:386-390.
[5]SHAO Y,SUN F,LI H.No-reference remote sensing image quality assessment method using visual properties.Journal of Tsinghua University,2013,53(4):550-555.(in Chinese)
[6]WANG Z,BOVIK A C,EVAN B L.Blind measurement of blocking artifacts in images.Proc.of the IEEE International Conference on Image Processing,2000:981-984.
[7]SAZZAD Z M P,KAWAYOKE Y,HORITA Y.No reference image quality assessment for JP2K based on spatial features.Signal Processing Image Communication,2008,23(4):257-268.
[8]HORITA Y,ARATA S,MURAI T.No-reference image quality assessment for JPEG/JP2K coding.Proc.of the IEEE Signal Processing Conference,2004:1301-1304.
[9]XIE F,PAN X,JIANG Z,et al.No-reference haze effects assessment for remote sensing images.Chinese Journal of Stereology&Image Analysis,2015,20(1):22-27.(in Chinese)
[10]WANG Z,BOVIK A C.Modern image quality assessment.New York:Morgan and Claypool Publishing Company,2006.
[11]MOORTHY A K,BOVIK A C.A two-step framework for constructing blind image quality indices.IEEE Signal Processing Letters,2010,17(5):513-516.
[12]LI C,BOVIK A C,WU X.Blind image quality assessment using a general regression neural network.IEEE Trans.on Neural Networks,2011,22(5):793-799.
[13]MOORTHY A K,BOVIK A C.Blind image quality assessment:from natural scene statistics to perceptual quality.IEEETrans.on Image Processing:A Publication of the IEEE Signal Processing Society,2011,20(12):3350-3364.
[14]SAAD M A,BOVIK A C,CHARRIER C.DCT statistics model-based blind image quality assessment.Proc.of the IEEE 18th International Conference on Image Processing,2011,9:3154-3157.
[15]SAAD M A,BOVIK A C,CHARRIER C.Blind image quality assessment:a natural scene statistics approach in the DCTdomain.IEEE Trans.on Image Processing:A Publication of the IEEE Signal Processing Society,2012,21(8):3339-3352.
[16]MITTAL A,SOUNDARARAJAN R,BOVIK A C.Making a“completely blind”image quality analyzer.IEEE Signal Processing Letters,2013,20(3):209-212.
[17]MITTAL A,MOORTHY A K,BOVIK A C.No-reference image quality assessment in the spatial domain.IEEE Trans.on Image Processing:A Publication of the IEEE Signal Processing Society,2012,21(12):4695-4708.
[18]MITTAL A,MOORTHY A K,BOVIK A C.Blind/referenceless image spatial quality evaluator.Proc.of the IEEE 45th Asilornar Conference on Signals,Systems and Computers,2011:723-727.
[19]LIU L,LIU B,HUANG H,et al.No-reference image quality assessment based on spatial and spectral entropies.Signal Processing Image Communication,2014,29(8):856-863.
[20]LI Q,LIN W,FANG Y.No-reference image quality assessment based on high order derivatives.Proc.of the IEEE International Conference on Multimedia and Expo,2016:1-6.
[21]LI Q,LIN W,XU J,et al.Blind image quality assessment using statistical structural and luminance features.IEEE Trans.on Multimedia,2016,18(12):2457-2469.
[22]LIU L,HUA Y,ZHAO Q,et al.Blind image quality assessment by relative gradient statistics and adaboosting neural network.Signal Processing Image Communication,2016,40(C):1-15.
[23]RUDERMAN D L.The statistics of natural images.Network Computation in Neural Systems,2009,5(4):517-548.
[24]WANG Z,SIMONCELLI E P,BOVIK A C.Multiscale structural similarity for image quality assessment.Proc.of the 37th IEEE Asilomar Conference on Signals,Systems&Computers,2003,2:1398-1402.
[25]SRIVASTAVA A,LEE A B,SIMONCELLI E P,et al.On advances in statistical modeling of natural images.Journal of Mathematical Imaging&Vision,2003,18(1):17-33.
[26]SHEIKH H R,WANG Z,BOVIK A C,et al.Image and video quality assessment research at LIVE 2004.http://live.ece.utexas.edu/research/quality.
[27]SHEIKH H R,SABIR M F,BOVIK A C.A statistical evaluation of recent full reference image quality assessment algorithms.IEEE Trans.on Image Processing,2006,15(11):3440-3451.
[28]SHARIFI K,LEON-GARCIA A.Estimation of shape parameter for generalized Gaussian distributions in subband decompositions of video.IEEE Trans.on Circuits&Systems for Video Technology,1995,5(1):52-56.
[29]ZHOU Y L,CHEN F,SUN B.Identification method of gas-liquid two-phase flow pattern based on image wavelet packet information entropy and genetic neural network.Nuclear Power Engineering,2008,29(1):115-120.(in Chinese)
[30]ZHANG Z.Digital image processing and machine vision.Beijing:The Posts and Telecommunications Press,2010.(in Chinese)
[31]CAO Y,MIAO Q G,LIU J C,et al.Advance and prospects of adaboost algorithm.Acta Automatica Sinica,2013,39(6):745-758.(in Chinese)
[32]WU J L,ZHANG B H,WANG K.Application of Adaboostbased BP neural network for short-term wind speed forecast.Power System Technology,2012,36(9):221-225.
[33]LI S,XIE Y L,WANG W X.Application of Ada Boost BP neural network in prediction of railway freight volumes.Computer Engineering and Applications,2012,48(6):233-234.(in Chinese)
[34]PONOMA N,PONOMARENKO N,JIN L,et al.Image database TID2013:peculiarities,results and perspectives.Signal Processing:Image Communication,2015,30:57-77.