Enterprise Attendance And Access Control System Design Based On Face Recognition

The Information Transformation of the personnel management is one of the most important revolution to the modern enterprise,the access control system based on face recognition has a huge development in recent yeas,but there are lots of weakness and defect in practical application,while the huge development of the deep learning algorithem brings a qualitative leap to the face recognition system.This thesis is builded on deep learning face recognition algorithem,aimed at the practical application of enterprise access control system,designed an access control system applicable to small and medium-sized companies.The thesis concentrates on four questions: the algorithem design and optimization for face detection and recognition;the model design,training and optimization based on Tensorflow;the record and storage of the model and attendance log;the design and realization of the GUI for system operation.The thesis describe and discuss the access control system from the following four aspects: the history of face recognition system and attendance system,basic technology,engineering design and results summarized.At first,the thesis introduces the research background and significance of the access control system,describes the research history and present situation of face recognition system.Through the presentation of the history of the access control system and the face recognition research experience,the thesis elaborates the basic concept,the idea,the form and the development direction in the future of the products and technologies,meanwhile expound the innovation and main contribution of the thesis according to the fact.Then the thesis mainly elaborates the basic theory,the application architecture and design framework of face recognition attendance system,the six common training model framework,six common deep learning network,and the basis of GUI software realization and comparison.Then the thesis introduces the design,implementation,and optimization of face recognition of attendance system.The main focus here is on algorithm implementation based on tensorflow,the design optimization of system modular and GUI for system operation.This chapter gave a detailed algorithm parsing and code parsing to the HOG+VGG algorithm,as well as a detailed illustration and code examples,which show the the engineering implementation of the system in all respects.Finally,the thesis darwn conclusions and analysis on the design idea,the actual results and the actual situationbased on multiple views,as well as claimed an expectation for the reasonable and practical direction for the later research and product development.