摘要
提出了一种融合局部二值模式(Local Binary Pattern,LBP)和BP神经网络的人脸表情识别方法。通过把人脸表情图像划分成若干区域,分别提取人脸子区域的LBP特征,然后把各个区域的LBP特征串联成一个完整的特征向量,该特征向量就可以表征该表情图像。构造BP神经网络人脸表情识别模型,通过在Cohn-Kanade人脸表情库的实验,验证了该方法的鲁棒性性和可行性。
This paper presents a local binary pattern（Local Binary Pattern,LBP） and BP neural network facial expression recognition method.By dividing the facial expression image into several regions,the LBP features of the face sub-region are extracted respectively,and then the LBP features of each region are concatenated into a complete eigenvector,which can characterize the expression image. Constructing BP neural network facial expression recognition model,the robustness and feasibility of the method is verified through the experiments in Cohn-Kanade facial expression database.
出处
《工业控制计算机》
2018年第5期69-70,共2页
Industrial Control Computer
