摘要
结合粒子群优化算法,提出了一种对人脸进行精确检测与分割的方法,以提高表情识别的准确率。通过肤色分割得到大致的人脸轮廓;通过粒子群优化检测到人脸的精确位置。利用精确的人脸进行PCA降维和特征提取后可进行表情识别。实验结果表明,此方法很好的排除了干扰因素,提高了表情识别的准确率,尤其对于某些单表情效果极佳。
A precise face detection and segmentation method combining with the particle swarm optimization is proposed. The aim is to improve the accuracy of a kind of special facial expression recognition problem. The method uses skin color segmentation to roughly obtain the face contour. The accurate face position is detected by particle swarm optimization. Then facial expression recognition is processed using precise face for PCA dimensionality reduction and feature extraction. Experimental results show that this method can eliminate the interference factor, and improve the accuracy of expression recognition, especially suitable for some single expression recognition.
出处
《大连民族学院学报》
CAS
2013年第3期305-308,336,共5页
Journal of Dalian Nationalities University
基金
国家自然科学基金(61040054)
中央高校基本科研业务费专项资金资助项目(DC120101081
DC120101082)
国家民委科研项目(12DLZ010)
关键词
范式表情识别
人脸检测
肤色模型
粒子群优化
主成分分析
facial expression recognition
face detection
skin color model
particle swarm optimization
principal component analysis