HVS模型及其在静止图象压缩质量评价中的应用

图象质量尺度是最优化图象压缩算法参数和提高图象质量的重要依据 .建立在人类视觉模型 (HVS)基础之上的感知质量尺度作为主客观联系的桥梁 ,能有效地反映出人对图象失真在视觉上的感知 .近年来 ,有许多研究者借助人类视觉系统研究中的最新成果 ,深入分析了与图象质量密切相关的视觉感知特性 ,提出了大量效果不错的静止图象压缩感知质量尺度 ,并对视觉感知特性在图象质量尺度中的应用方法进行了较全面的综述 ,揭示了影响其图象质量预测准确性、鲁棒性的主要因素 ,给出了该领域的最新研究成果和未来发展方向 .

Human Visual System Models for the Quality Assessment of the Still Image Compression

Image quality metric could be utilized to optimize the image compression algorithm and improve the image quality. Perceptual quality metric, based on the foundation of human visual system (HVS), can be used as a kind of closed connection between subjective assessments and objective assessments and reflects the human's visual sense to the image distortion. In recent years, the visual perception characteristics in relation to image quality closely has been researched thoroughly by means of some new progress in HVS, and a lot of perceptual quality metrics on the still image compression which all sound effectively have been proposed. An almost comprehensive survey of the different application of the visual perception characteristics in perceptual quality metric is presented in this paper, the important factors to predict the image quality accurately and robustly is given and the research achievement is summarized in this field. However, the development of computational HVS models is still in its infancy, and many issues remain to be investigated and solved. First of all, more psychophysical experiments need to be done with natural images for the modeling of more complex phenomena that occur in natural images. Secondly, more psychophysical experiments focus on measurements at the supra threshold also need to be done because quality metrics and compression are often applied above threshold. Finally, HVS models would be expressed by analysis functions with the aid of latest mathematic pay off in order that the general image quality metrics are developed.