目前,视频业务在网络服务中所占比重越来越大。对视频业务质量进行评估有助于服务提供商改善用户体验,具有重要研究意义。在客观视频质量评价方法中,基于结构相似性的视频评价算法(SSIM)性能良好,计算简单,因此得到了广泛应用与研究。然而这种方法在评价某些模糊失真时性能不佳。针对于此,本文提出一种考虑人眼视觉特性的视频质量评价算法。本文根据人眼视觉特性(HVS),在结构相似性算法中集成了对比敏感度,多信道结构,以及边缘效应等特性。实验结果表明,与结构相似度算法相比,本文提出的方法与主观视频质量评价具有更好的相关性和更高的预测精度,也更符合人眼的主观感受。特别是对于视频压缩失真的评价性能具有明显的提高。%Currently, the video network services business in the proportion is growing. Assessing the quality of the video s help service providers improve the user experience. So the search has important significance. Because of its good performance and simple calculation, the image quality assessment algorithm called structural similarity (SSIM) has been widely concerned. However, this method has poor performance in the evaluation of some fuzzy distortion. In light of this, this paper proposes a video quality assessment algorithm considering human visual characteristics. According to the human visual system (HVS), this paper integrates the contrast sensitivity, multi-channel structure, edge effects and other features into the structural similarity algorithm. Experimental results show that this method is better correlated with subjective video quality evaluation ,more accurate and more consistent with human's subjective feelings, especially for compression distortion.
展开▼