基于视觉感知的LCDs运动图像边缘模糊的评价方法

Evaluation method of edge sharpness for motion images on LCDs based on perceptual experiment

基于LCD运动图像锐度的视觉感知实验,采用线性回归的方法,得到人眼感知到的运动图像边缘锐度参量Zscore的计算表达式,提出了利用与边界斜率和过驱动相关的2个参数来描述运动图像边缘模糊程度的量化公式.并将计算预测值和视觉感知的实验值进行比较.结果表明,二者具有较高的相关性(R2≥0.889 2),该量化公式可以很好地用于计算感知运动图像的锐度,并改善采用现有评测运动模糊的标准方法时测量评价结果和人眼感知到的运动图像的边缘模糊程度不一致的现象.

A new calculation method and expression for the perceived sharpness metric in LCDs（liquid crystal displays） is proposed.Based on visual perception experiment,two new evaluation parameters Cslope and OD（over-drive）,related to the edge slope and over/under-shoot part of luminance edge profile respectively,are put forward to predict the perceived sharpness.Experiments are conducted to verify the feasibility and effectiveness of this predicted perceived sharpness formula.The results show a high correlation between the metric predictions and perceptual experiment（R2≥0.8892）.The method proposed can improve the limitation of existing standards in quantitative evaluation of complicated motion edge blur.