基于网络摄像头的活体人脸识别系统设计

Design of Living Face Recognition System Based on Web Camera

针对当前监控系统平台体积大、功能单一、造价高等问题,文中基于海思3518E平台设计了一套在线活体人脸识别系统。该系统包含图像预处理、图像获取、人脸检测、人脸活体检测、人脸识别共5个部分,并针对传统算法在嵌入式平台的不足进行了改进与优化。系统通过摄像头采集的人脸图像对图像做预处理,为后续图像特征提取提供保障,利用扩展Haar特征训练分类器,并使用Adaboost算法级联分类器进行人脸检测,将检测到的人脸利用HSV和Ycbcr多色彩空间下提取的COALBP和LPQ融合特征训练SVM模型,并进行活体人脸检测。最后,对人脸图像分块提取LBP特征进行人脸识别,将识别结果通过微信小程序显示。实验结果表明,基于海思网络摄像头人脸识别系统的可行性,具有一定的实用价值。

Aiming at the problems of large size,single function and high cost of the current monitoring system platform,this paper designs an online living face recognition system based on the HiSilicon 3518E platform.The system includes five parts:image pre-processing,image acquisition,face detection,face live detection,and face recognition.It also improved and optimized the deficiencies of traditional algorithms in embedded platforms.The system collected the face image through the camera and preprocessed the image to provided guarantee for subsequent image feature extraction.The extended Haar feature was used to train the classifier and the Adaboost algorithm was used to cascade face detection.The detected face used HSV and Ycbcr multi-color space the COALBP and LPQ fusion features extracted below were used to train the SVM model,perform live face detection,and finally extract LBP features from the face image to perform face recognition,and the recognition results were displayed through the WeChat applet.The experimental results showed that the feasibility of the face recognition system based on the HiSilicon web camera had certain practical value.