Abstract: The law of universal gravitation is introduced to face recognition. A model of identification based on the law is established. Conventional method of Principal Component Analysis(PCA) is performed to extract the features of face image. The similarity between two face images and the expression of gravitation between each image and each class are formulated. The magnitude of the gravitation decides to which class the face image belongs. Experimental results obtained on ORL show that the model is feasible .

Key words: universal gravitation, Principal Component Analysis(PCA), feature extraction

摘要： 把万有引力定律引入到人脸识别中，构建基于万有引力定律的人脸识别模型。用传统的主成分分析方法对图像进行特征提取，建立图像与图像之间的相似度模型以及各个类别对图像潜在作用力的表达式，根据该作用力大小决定图像所属类别。在ORL人脸数据库上的实验表明，该模型有一定的可行性。

关键词: 万有引力, 主成分分析, 特征提取

CLC Number: 

• TP391