基于双目融合与竞争的无参考立体图像质量评价方法

No-reference Stereoscopic Image Quality Assessment Based on Binocular Fusion and Binocular Rivalry

根据人眼对立体图像的感知过程,提出了一种基于双目融合和竞争特性的无参考立体图像质量评价方法.首先将左右视点图像进行融合,对得到的独眼图进行Gabor特征提取;然后对左右视点图像的绝对差值图提取特征;最后将独眼图特征和绝对差值图特征融合得到立体图像特征集,通过支持向量回归预测得到客观值.采用该方法对LIVE立体图像数据库进行评价,Pearson线性相关系数(PLCC)和Spearman等级相关系数(SROCC)均在0.94左右,优于其他参与对比的质量评价方法.表明该方法符合人眼视觉特性,能够很好地描述人眼感知特性.

Stereoscopic image quality assessment （SIQA） plays an important role in stereo video system applications. According to the processing of binocular perception in watching stereo image, a novel no-reference SIQA based on binocular fusion and binocular rivalry is proposed. In this method, firstly, Gabor filtering features are extracted from the binocular fusion image formed by both left and right view images. Secondly, the natural features are abstracted from absolute difference map. Finally, two parts of features are incorporated to form stereoscopic image feature information, and support vector regression is used to predict the objective scores by establishing the relationship between the stereoscopic image feature information and the subjective scores. By applying the proposed method to the test stereoscopic image database, it demonstrates that correlation coefficients are all about 0.94, better than the other representative image quality assessment methods, suggesting that the proposed method has better correlation with human subjective visual quality evaluation compared with some existing methods.