期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2010
卷号：XXXVIII - Part 2
页码：124-128
出版社：Copernicus Publications
摘要：Facial expression is one of the major features of facial recognition in recent years, and it has become a hotspot. In this paper, we present a novel method of facial recognition based on cloud model, in combination with the traditional Facial expression system. Firstly, we carry out the transformation from images into grids with M by N, where M and N denote the actual image positioning of the grid. Each grid is a gray value (0-255) and the grids stand for the data from data points to data sets based on cloud model. Secondly, we do data pre-processing for the original facial expressions of input images. Cloud droplets image can be obtained as the input of backward cloud generator in order to extract the three numerical characteristics, that is, Ex, En and He. With these three characteristics, facial expression can be realized. Finally, in order to demonstrate the feasibility of the presented method, we conduct a case study of facial expression recognition based on cloud model. The results show that the method is feasible and effective in facial expression recognition
