沈丽丽 彭科

摘 要：立体图像质量评价在立体图像处理领域中应用广泛.基于小波包分解的精细分辨率，提出了一种全新的无参考立体图像质量评价算法.选取合值图和差值图作为融合图来评估立体图像，首先，对立体图像对进行小波包分解，基于双眼竞争和双眼抑制原理，将分解后的左右视图进行融合得到合值图和差值图.然后，分别在融合图上提取自然场景统计（NSS）特征和信息熵；另外，考虑到左右视图之间的内在相关联系，提取结构相似度特征.最后，运用支持向量回归（SVR）来建立感知特征和主观分数模型并预测得到客观评价分数.采用该算法在LIVE 3D立体图像数据库上进行测试，实验结果表明，该算法与人眼主观评价结果一致性较高，优于当前主流的立体图像质量评价算法，符合人眼视觉感知特性.

关键词：立体图像质量评价；小波包；无参考；结构相似度

中图分类号：TP391 文献标志码：A

Abstract：Stereoscopic image quality evaluation is widely employed in the field of stereo image processing. Based on accurate resolution of waveletpacket， a novel algorithm for noreference stereoscopic image quality assessment is proposed. It evaluates stereo image quality via fusion map， which consists of cyclopean map and difference map. First， the stereo image pair is decomposed by waveletpacket， and afterwards the decomposed left and right views are fused to obtain cyclopean map and difference map based on the principle of binocular rivalry and binocular suppression. Then， Natural Scene Statistics （NSS） features and information entropy are extracted on the fusion map； besides， the structural similarity feature is extracted by taking into account the internal relations between the left and right views. Finally， the Support Vector Regression （SVR） is used to establish model between the perception features and subjective scores， which can predict the objective evaluation score. The experimental results on LIVE 3D image databases show that the proposed algorithm has high consistency with the subjective evaluation results， and it outperforms stateoftheart stereoscopic image quality assessment algorithms and is in accordance with the human visual perception characteristics.

Key words： stereoscopic image quality assessment（SIQA）； waveletpacket； noreference（NR）； structural similarity

近年來，立体电视、立体游戏等立体技术已逐渐深入到人们的日常生活当中，为人们带来了极佳的视觉体验和心理震撼.相比于二维平面图像，立体图像可以使人们身临其境地感受到立体感及临场感，逐渐地成为信息技术研究的主流方向[1].然而，在3D立体图像成像过程中，图像信息采集、压缩、传输、解码、渲染等会存在各种各样的失真，失真程度非常严重的立体图像会使人观看不舒服，甚至出现疲劳、头晕等现象.因此，对立体图像进行质量评估非……