基于RPSO_SVM模型的年龄组识别  

Age Group Recognition Based on RPSO_SVM Model

作　　者：邵定琴 张乾[1] 岳诗琴 白金华 苏江涛 SHAO Ding-qin;ZHANG Qian;YUE Shi-qin;BAI Jin-hua;SU Jiang-tao(Guizhou Minzu University,Guiyang 550025,China)

机构地区：[1]贵州民族大学,贵州贵阳550025

出　　处：《电脑知识与技术》2021年第25期20-23,共4页Computer Knowledge and Technology

基　　金：国家自然科学基金项目(No.61802082,No.61263034,No.61762020)。

摘　　要：不同年龄段群体在应用市场上的需求存在明显差异性,年龄组的研究具有重要现实意义。通过改进粒子群算法对支持向量机参数进行优化(RPSO_SVM),用于青年人、中年人、老年人等群体的年龄组识别,首先,使用主动形状模型提取人脸图像中68个特征关键点,然后,使用改进的RPSO_SVM模型对人脸图像进行年龄组识别,并在FG-NET数据集上进行实验验证,结果表明该方法对不同年龄群体的识别率较好。There are obvious differences for different age groups in the application market,it has important practical significance to study age groups.This paper optimizes the Support Vector Machine parameters by improving the Particle Swarm Optimization(RP⁃SO_SVM),which is used for the age group recognition of young people,middle-aged people,and old people.First,Active Shape Model is used to extract 68 feature key points in the face image,and then the improvement RPSO_SVM algorithm is used to identi⁃fy the age groups of face images,and verified by experiments on the FG-NET dataset.The results show that the method has a better recognition rate for different age groups.

关 键 词：粒子群算法 支持向量机 年龄组识别 主动形状模型 RPSO_SVM模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]