Research And Implementation Of Target Detection And Tracking Algorithm Based On Face Recognition

Under the background of the new era,because of people’s sharp judgment on the development of things,new concepts and demands such as intelligent city,intelligent home and intelligent monitoring have been put forward one after another,and all of these are based on artificial intelligence technology as the core.It is a new field of computer technology,interconnection communication and automatic control theory.It has developed rapidly in recent years,and has turned from experimental research and development to commercial promotion.At present,it has been used in machine vision,face recognition,and so on.Intelligent search,language and image understanding are widely used.The new intelligent thinking is also used in security system construction.It can be used in the construction of security system,which can form the face,traffic flow by implanting an efficient algorithm with artificial intelligence into the monitoring equipment.The integrated system of target extraction,detection and recognition,and behavior understanding,such as license plate information,will be the symbol of the change from security system to intelligence in the future.This paper is the design of the proposed intelligent security Home Furnishing environment solution.Because the ARM processor implementation of high efficiency,low cost,performance and economic advantages of miniaturization,and easy to transplant Android operating system,to facilitate the development of the mobile version of the interface software,the choice of Imx6 development board as the hardware foundation.According to the feature that the monitoring emphasizes the function of human body target detection,through the comprehensive analysis of various image background detection,feature extraction,target recognition and tracking algorithm,ViBe algorithm is selected to quickly separate foreground targets and extract facialfeatures.The strong classifier based on Haar facial features is generated by using AdaBoost algorithm,and the Kalman filter algorithm is used to solve the problem that the monitoring picture is affected by indoor illumination,and the face occlusion and other factors lead to the reduction of face recognition rate.Finally,the basic function of smart home security is realized by writing programs on QT to call functions in OpenCV,and design the interface of monitoring software.The whole system designed on the basis of cortex-A9 processor hardware which based on Linux kernel.It mainly includes video information collection,human body target detection and recognition,tracking and intrusion alarm,etc.Practice has proved that it can meet the basic requirements of real-time and accuracy.