基于双树复小波的无参考立体图像质量评价
|
摘要
随着3D技术的不断发展,立体图像的使用领域越来越广泛,同时人们对图像的清晰度要求越来越高,因此,立体图像的质量评价成为关注点,基于此,提出了一种基于双树复小波变换的立体图像质量评价算法。使用双树复小波变换对立体图像的左、右视图进行处理,生成纹理结构图像,且根据最小能量误差的原理,获取左右视图的视差图;对纹理结构图像和视差图提取非对称广义高斯分布模型的参数、梯度幅值、相对梯度方向方差和奇异值曲线与坐标轴的面积等特征;使用AdaBoosting BP神经网络,进行训练和预测立体图像的质量得分。在LIVE立体图像数据库上的实验结果表明,新方法预测得分与主观得分有较好的一致性,获得了比较好的实验结果。
With the rapid development of 3D technology, stereoscopic images are widely used in many fields. At the same time, it is expected that high-definition images should be provided. So image quality assessment for stereoscopic images has been the concern and a no-reference image quality assessment algorithm for stereoscopic images via Dual-Tree Complex Wavelet Transform(DT-CWT)is proposed. Firstly, the left and the right view images of a stereoscopic image need to be processed in the way of DT-CWT to generate the textural structure image. And the parallax image is obtained according to the principle of minimum energy error. Secondly, features from the textural structure image and the parallax image are extracted, including the parameters of Asymmetric Generalized Gaussian Distribution(AGGD), variance of Gradient Magnitude(GM)and Relative gradient Orientation(RO), the area of Singular Value Decomposition(SVD)curve and the coordinate axis. Lastly, it trains AdaBoosting Back-Propagation(BP)neural network by utilizing these features and predicts the quality scores of stereoscopic images. The experiment on LIVE 3D Image Quality Database demonstrates that the predicted scores by the new method have high correlation with mean opinion scores and a good evaluation result is obtained.
With the rapid development of 3D technology, stereoscopic images are widely used in many fields. At the same time, it is expected that high-definition images should be provided. So image quality assessment for stereoscopic images has been the concern and a no-reference image quality assessment algorithm for stereoscopic images via Dual-Tree Complex Wavelet Transform(DT-CWT)is proposed. Firstly, the left and the right view images of a stereoscopic image need to be processed in the way of DT-CWT to generate the textural structure image. And the parallax image is obtained according to the principle of minimum energy error. Secondly, features from the textural structure image and the parallax image are extracted, including the parameters of Asymmetric Generalized Gaussian Distribution(AGGD), variance of Gradient Magnitude(GM)and Relative gradient Orientation(RO), the area of Singular Value Decomposition(SVD)curve and the coordinate axis. Lastly, it trains AdaBoosting Back-Propagation(BP)neural network by utilizing these features and predicts the quality scores of stereoscopic images. The experiment on LIVE 3D Image Quality Database demonstrates that the predicted scores by the new method have high correlation with mean opinion scores and a good evaluation result is obtained.
引文
[1]Chen M J,Cormack L K,Bovik A C.No-reference quality assessment of natural stereopairs[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2013,22(9):3379-3391.
[2]You J,Xing L,Perkis A,et al.Perceptual quality assessment for stereoscopic images based on 2D image quality metrics and disparity analysis[C]//International Workshop on Video Processing and Quality Metrics for Consumer Electronics,2010.
[3]Benoit A,Callet P L,Campisi P,et al.Quality assessment of stereoscopic images[J].EURASIP Journal on Image&Video Processing,2008(1):1-13.
[4]Xing L,You J,Ebrahimi T,et al.A perceptual quality metric for stereoscopic crosstalk perception[C]//IEEE International Conference on Image Processing,2010:4033-4036.
[5]Yang J,Hou C,Zhou Y,et al.Objective quality assessment method of stereo images[C]//3DTV Conference:the True Vision-Capture,Transmission and Display of 3d Video,2009:1-4.
[6]Bensalma R,Larabi M C.A perceptual metric for stereoscopic image quality assessment based on the binocular energy[J].Multidimensional Systems&Signal Processing,2013,24(2):281-316.
[7]Shao F,Lin W,Gu S,et al.Perceptual full-reference quality assessment of stereoscopic images by considering binocular visual characteristics[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2013,22(5):1940-1953.
[8]Zhang W,Ma L,Ma L,et al.Learning structure of stereoscopic image for no-reference quality assessment with convolutional neural network[J].Pattern Recognition,2016,59(C):176-187.
[9]Shao F,Lin W,Wang S,et al.Learning receptive fields and quality lookups for blind quality assessment of stereoscopic images[J].IEEE Transactions on Cybernetics,2016,46(3):730.
[10]Zhou W,Yu L,Qiu W,et al.Utilizing binocular vision to facilitate completely blind 3D image quality measurement[J].Signal Processing,2016,129:130-136.
[11]Ryu S,Sohn K.No-reference quality assessment for stereoscopic images based on binocular quality perception[J].IEEE Transactions on Circuits&Systems for Video Technology,2014,24(4):591-602.
[12]Zhong X,Li C,Zhang W,et al.No-reference image quality assessment using dual-tree complex wavelet transform[C]//International Congress on Image and Signal Processing,2015:596-601.
[13]田维军,邵枫,蒋刚毅,等.基于深度学习的无参考立体图像质量评价[J].计算机辅助设计与图形学学报,2016,28(6):968-975.
[14]Wang Z,Simoncelli E P,Bovik A C.Multiscale structural similarity for image quality assessment[C]//Conference Record of the Thirty-Seventh Asilomar Conference on Signals,Systems and Computers,2004:1398-1402.
[15]Kingsbury N.A dual-tree complex wavelet transform with improved orthogonality and symmetry properties[C]//International Conference on Image Processing,2000:375-378.
[16]Roth J P,Scott Lovell?A,Mayer J M.Intrinsic barriers for electron and hydrogen atom transfer reactions of biomimetic iron complexes[J].Journal of the American Chemical Society,2000,122(23):5486-5498.
[17]吴一全,尹丹艳,纪守新.基于双树复数小波和SVR的红外小目标检测[J].仪器仪表学报,2010,31(8):1834-1839.
[18]周俊明,郁梅,蒋刚毅,等.利用奇异值分解法的立体图像客观质量评价模型[J].计算机辅助设计与图形学学报,2011,23(5):870-877.
[19]Wang L,Duan F Q,Liang C.A global optimal algorithm for camera calibration with one-dimensional objects[C]//International Conference on Human-Computer Interaction.Berlin,Heidelberg:Springer,2011:660-669.
[20]Lasmar N E,Stitou Y,Berthoumieu Y.Multiscale skewed heavy tailed model for texture analysis[C]//IEEE International Conference on Image Processing,2009:2281-2284.
[21]Heidary K,Caulfield H J.Margin-setting with hyperellipsoidal surfaces[J].Optical Memory&Neural Networks,2010,19(2):97-109.
[2]You J,Xing L,Perkis A,et al.Perceptual quality assessment for stereoscopic images based on 2D image quality metrics and disparity analysis[C]//International Workshop on Video Processing and Quality Metrics for Consumer Electronics,2010.
[3]Benoit A,Callet P L,Campisi P,et al.Quality assessment of stereoscopic images[J].EURASIP Journal on Image&Video Processing,2008(1):1-13.
[4]Xing L,You J,Ebrahimi T,et al.A perceptual quality metric for stereoscopic crosstalk perception[C]//IEEE International Conference on Image Processing,2010:4033-4036.
[5]Yang J,Hou C,Zhou Y,et al.Objective quality assessment method of stereo images[C]//3DTV Conference:the True Vision-Capture,Transmission and Display of 3d Video,2009:1-4.
[6]Bensalma R,Larabi M C.A perceptual metric for stereoscopic image quality assessment based on the binocular energy[J].Multidimensional Systems&Signal Processing,2013,24(2):281-316.
[7]Shao F,Lin W,Gu S,et al.Perceptual full-reference quality assessment of stereoscopic images by considering binocular visual characteristics[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2013,22(5):1940-1953.
[8]Zhang W,Ma L,Ma L,et al.Learning structure of stereoscopic image for no-reference quality assessment with convolutional neural network[J].Pattern Recognition,2016,59(C):176-187.
[9]Shao F,Lin W,Wang S,et al.Learning receptive fields and quality lookups for blind quality assessment of stereoscopic images[J].IEEE Transactions on Cybernetics,2016,46(3):730.
[10]Zhou W,Yu L,Qiu W,et al.Utilizing binocular vision to facilitate completely blind 3D image quality measurement[J].Signal Processing,2016,129:130-136.
[11]Ryu S,Sohn K.No-reference quality assessment for stereoscopic images based on binocular quality perception[J].IEEE Transactions on Circuits&Systems for Video Technology,2014,24(4):591-602.
[12]Zhong X,Li C,Zhang W,et al.No-reference image quality assessment using dual-tree complex wavelet transform[C]//International Congress on Image and Signal Processing,2015:596-601.
[13]田维军,邵枫,蒋刚毅,等.基于深度学习的无参考立体图像质量评价[J].计算机辅助设计与图形学学报,2016,28(6):968-975.
[14]Wang Z,Simoncelli E P,Bovik A C.Multiscale structural similarity for image quality assessment[C]//Conference Record of the Thirty-Seventh Asilomar Conference on Signals,Systems and Computers,2004:1398-1402.
[15]Kingsbury N.A dual-tree complex wavelet transform with improved orthogonality and symmetry properties[C]//International Conference on Image Processing,2000:375-378.
[16]Roth J P,Scott Lovell?A,Mayer J M.Intrinsic barriers for electron and hydrogen atom transfer reactions of biomimetic iron complexes[J].Journal of the American Chemical Society,2000,122(23):5486-5498.
[17]吴一全,尹丹艳,纪守新.基于双树复数小波和SVR的红外小目标检测[J].仪器仪表学报,2010,31(8):1834-1839.
[18]周俊明,郁梅,蒋刚毅,等.利用奇异值分解法的立体图像客观质量评价模型[J].计算机辅助设计与图形学学报,2011,23(5):870-877.
[19]Wang L,Duan F Q,Liang C.A global optimal algorithm for camera calibration with one-dimensional objects[C]//International Conference on Human-Computer Interaction.Berlin,Heidelberg:Springer,2011:660-669.
[20]Lasmar N E,Stitou Y,Berthoumieu Y.Multiscale skewed heavy tailed model for texture analysis[C]//IEEE International Conference on Image Processing,2009:2281-2284.
[21]Heidary K,Caulfield H J.Margin-setting with hyperellipsoidal surfaces[J].Optical Memory&Neural Networks,2010,19(2):97-109.