基于蜂群算法的人脸表情特征选择方法研究  

Feature Selection Method for Face Expression Based on ABC Algorithm

作　　者：高凤娇[1] 吴冈[1] 费磊[1] 杨洋[1] 

机构地区：[1]黑龙江省科学院自动化研究所,黑龙江哈尔滨150090

出　　处：《自动化技术与应用》2013年第3期74-76,80,共4页Techniques of Automation and Applications

摘　　要：特征选择技术是模式识别系统的一个关键步骤,在人脸表情识别分析过程中特征选择步骤也尤为重要。在众多的特征中选择合适的特征组合对人脸表情的识别分析影响很大。本文提出基于蜂群算法的人脸表情特征选择方法,利用蜂群算法寻找到适当的人脸表情特征组合,减少全部特征下识别过程中的计算量,同时提高识别率。并且通过实验证明了本文方法的可行性和优越性。Feature selection technology is one of the key steps processing of face expression recognition analysis. It of pattern recognition system. It＇s procedure is also vital in the affects facial expression recognition analysis choosing the right combination of facial expression ones in many of the characteristics. This paper puts forward the method based on Artificial Bee Colony algorithm for face expression feature selection, using Artificial Bee Colony algorithm to find the appropriate facial expression characteristic combinations, to reduce calculation quantity under using all the characteristics, at the same time improve the recognition rate. Experiments prove the feasibility and su^erioritv of this method

关 键 词：蜂群算法 人脸表情 特征选择 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]