\r沈丽丽，王 莹\r
\r
AuthorsHTML:\r沈丽丽，王 莹\r
\r
AuthorsListE:\rShen Lili，Wang Ying\r
\r
AuthorsHTMLE:\rShen Lili，Wang Ying\r
\r
Unit:\r天津大学电气自动化与信息工程学院，天津 300072\r
\r
Unit_EngLish:\rSchool of Electrical and Information Engineering，Tianjin University，Tianjin 300072，China\r
\r
Abstract_Chinese:\r\r针对非对称失真立体图像，提出了一种基于奇异值分解的无参考评价算法．该方法首先考虑人眼对空间频率变化敏感的特性和双目融合特性，对立体图像进行\rGabor\r滤波，基于奇异值分解的融合策略生成融合图．然后，采用亮度加权直方图的局部二值模式算法分别对融合图、左右子图像提取特征，并将左右子图像的特征向量融合、采用欧几里得距离和夹角余弦进行向量之间的比较；为度量非对称失真差异，利用图像相似度算法计算左右子图像之间的相似性．最后，将融合图的特征向量、子图像的融合及比较特征向量、子图像的相似度特征向量级联，利用支持向量回归\r(\rSVR\r)\r算法完成特征到主观质量分数的回归映射．在\rLIVE 3D \rⅡ\r、\rWaterloo\r-\rIVC \rⅠ\r和\rWaterloo\r-\rIVC \rⅡ\r立体图像库上对本算法进行测试．实验结果表明，本算法性能良好，优于目前主流的立体图像质量评价算法．\r\r
\r
Abstract_English:\r\rFor asymmetrically distorted stereoscopic images\r，\ra no-reference evaluation algorithm based on singular value decomposition is proposed\r．\rFirst\r，\rconsidering visual sensitivity to spatial frequency variation and binocular fusion\r，\rGabor filtering was performed on the stereoscopic image\r，\rand a fusion strategy based on singular value decomposition was proposed to generate a cyclopean image of the left and right sub-image pair\r．\rThen\r，\rthe proposed luminance-weighted histogram local binary pattern metric was used to extract features of the cyclopean image and the left and right sub-images\r．\rFurthermore\r，\rfeature fusion and comparison were conducted on the two feature vectors corresponding to the left and right sub-images\r，\rrespectively\r．\rEuclidean distance and cosine were used to implement the vector comparison\r．\rParticularly\r，\rto measure the difference between asymmetrically distorted sub-image pair\r，\rimage similarity metric was utilized to calculate the similarity between the left and right sub-image pair\r．\rFinally\r，\rfeature vector of the cyclopean image\r，\rthe fusion and comparison feature vectors\r，\rand the similarity feature vector were concatenated into a total feature vector\r，\rand regression mapping was performed from the feature vector to the subjective score using support vector regression\r．\rThe algorithm was tested on the LIVE 3D \rⅡ\r，\rWaterloo-IVC \rⅠ\r and Waterloo-IVC \rⅡ\r databases\r．\rThe experimental results show that the proposed algorithm has an outstanding performance and is superior to other state-of-the-art image quality assessment metrics\r．\r\r
\r
Keyword_Chinese:立体图像质量评价；非对称失真；奇异值分解；欧几里得距离；图像相似度\r

Keywords_English:stereoscopic image quality assessment；asymmetric distortion；singular value decomposition；Euclidean distance；image similarity\r


PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=6473