Design And Implementation Of Detection And Tracking System Based On Face Recognition

Face recognition is an advanced biometric identification technique.Its main purpose is that the computer application can automatically identify and verify a person from a digital image or a video frame.This technology has been widely used in information security,entrance control,certificate verification and other fields.Similarly,target tracking is one of the most important research in computer vision field,which has played an important role in real-time monitoring and traffic control.In recent years,with the development of artificial intelligence and robot technology,the interaction between human and robot has become more and more important.Therefore,the application of target tracking technology to robot has certain practical value.At the same time,face feature is the ideal biometric feature in identity authentication.Its uniqueness,noninvasiveness and easy to sample are the main advantages in the process of target tracking.In this paper,the system design is based on software and hardware,optimizes the face recognition algorithm,narrows the distances between classes,reduces the negative impact on errors of the acquisition process and improves face recognition accuracy.The main contents of this paper are as follows:Firstly,complete the overall framework of the system and the design of each module.With VS2010,QT creator,CMake,mysql5.6,navicat for mysql and OpenCV 2.4.6 computer vision library as the software platform,the design of detecting system based on face recognition includes the modules: algorithm module design,database module design,PTZ control module and interactive interface design.Secondly,to improve the face recognition algorithm,based on extracting face Gabor wavelet texture feature and the application of PCA on the principal component analysis of dimensionality reduction,improves the sample stage with K-Means clustering algorithm,illumination change or expression differences caused by the error between the target class.Lastly,with the camera,PTZ,computer as the hardware platform,the system will sent the target position to PTZ.This can make camera follow the target face.Experimental results show that the improved algorithm has a significant improvement in the accuracy of face recognition.Especially for the case of large errors in the process of collecting samples,it has good robustness.Finally,this paper discusses the significance and value of the system in real life.