论文部分内容阅读
针对现有Chernoff脸谱图(以下简称脸谱图)的绘制过程并未考虑变量分配对其最终表达能力的影响,本文尝试提出改进脸谱图的思路。首先通过人类对不同脸谱图物件(包括脸部器官和发型等)的识别敏感度进行降序排序;其次对变量的变异程度依方差、变异系数、极差同样进行降序排序;再次依强强结合的原则将变异程度高的变量分配到识别敏感度高的(变异程度低的变量则分配到识别敏感度低的)构件上,实现脸谱图的改进,增强脸谱图的表达能力;最后利用人体尺寸数据与树叶轮廓数据对几个方案进行比较,得出基于方差与极差的变量分配所得到的脸谱图明显优于现行基于变量随机分配的脸谱图的结论。
The drawing process of the existing Chernoff face chart (hereinafter referred to as the face chart) does not consider the influence of the variable assignment on its final expression ability. This article attempts to propose the idea of improving the face chart. First of all, the recognition sensitivity of different facial map objects (including facial organs and hairstyles, etc.) is sorted in descending order by humans. Second, the degree of variation of variables is also sorted in descending order according to variance, coefficient of variation and range difference; In principle, the variable with high degree of variability is allocated to the component with high sensitivity (low variability is assigned to the low sensitivity) to improve the facial map and enhance the expression ability of the facial map; Finally, the use of body size data Compared with the leaf contour data of several schemes, it is concluded that the face mapping obtained based on the variance and the bad variance assignment is obviously superior to the existing face mapping based on the random assignment of variables.