Design Of Intelligent Site Attendance System Based On Face Recognition

In recent years,the construction industry has put forward an innovative management method of the real name system of labor employment,which can effectively strengthen the legal and compliance management of labor employment on site,and effectively protect the rights and interests of migrant workers.When there is a labor dispute between migrant workers and labor companies,Accurate entry and exit records can provide true and strong evidence for the authorities' fair judgment.Labor companies can also effectively avoid labor disputes if they have the precise labor track of the migrant workers on the construction site.It is very important to accurately grasp the entry record of construction site operators to discover,eliminate and prevent the hidden trouble of labor dispute in labor management.Intelligent site attendance system of identification technology came into being.The intelligent site attendance system based on face recognition mainly consists of two parts,one is face recognition technology,the other is attendance system.And the user face information is the only condition to produce the system attendance record.This paper mainly includes face image preprocessing,face detection,face recognition and system design.(1)In the face image preprocessing,the weighted average method is used to process the image grayscale,and then the histogram equalization and median filter are used to process the image.It can effectively reduce the interference of light and dark caused by external factors such as illumination condition and hardware equipment.After the human eye is positioned,the face position is rotated and corrected,and the image size is normalized after the face is detected.It then shrinks or expands to the standard size required for the experiment.(2)The classical face detection methods,including feature-based face detection,template matching and statistic based face detection,are described in this paper.Then,the Adaboost face detection method is emphasized.This paper studies the theoretical foundation of Adaboost algorithm and the application of integral graph,and expounds the design of cascade classifier.In this paper,the basic idea and flow of Adaboost algorithm are presented.In this paper,the cascaded strong classifier formed by different weak classifiers is used by Adaboost algorithm for face detection.(3)In the detection part,based on the basic LBP algorithm,an algorithm based on block weighted LBP technology is proposed.The face image is divided into 5 sub-blocks and 3 sub-blocks.According to the different contribution degree of facial facial features in face recognition,the feature of face information is extracted with different weights.By training different sample numbers in ORL and YALE face database,the accuracy of facial recognition using traditional LBP methods,such as LBP method and weighted LBP method,is compared.Experiments show that block weighted LBP technology can effectively improve the accuracy of face recognition.(4)The whole system is designed in detail.The background database,identity collection and face database of the system are discussed in detail.At the same time,the user login of the system and the management of the working group are introduced in detail.The realization of the four modules of migrant workers management and site attendance management.Finally,the intelligent site attendance system based on face recognition is designed.