3D Face Recognition System Based On Kinect And Its Application In Medical Insurance Review

With the development of science and technology,face recognition technology is increasingly applied in railway station security check,residential access control system,mobile payment,hospitals and drug stores.The traditional 2D face recognition technology is mostly affected by external factors such as illumination,posture,occlusion,etc.,which leads to unrecognizable or incorrect recognition.There are many hidden dangers.3D face recognition has received more and more attention due to its advantages such as illumination,posture,robustness of makeup and richness of facial information.Firstly,this thesis uses a fast and low-cost Kinect 3D sensor to design and implement a 3D face recognition subsystem.The front-end of the system is based on the Vue.js framework,the back-end is based on the Spring framework,and the database uses MySQL.The system implements front-end and back-end separation,which can improve the development efficiency.Among them,the bone positioning technology used in face detection is not affected by light,which makes up for the deficiency of light factor in twodimensional images.Face recognition uses Keras framework to train twin neural network for predictive classification.Through data acquisition,face model generation,model training and face recognition function modules,face recognition is effectively realized.Secondly,a prototype of medical insurance review system based on 3D face recognition was designed and implemented.The prototype consists of three parts: front end,server back end and database management.The main functional modules of the system include patient registration,system login,patient recognition and database management.Among them,the patient recognition part calls the interface of the threedimensional face recognition subsystem,assists the medical insurance review system to complete the identify authentication,and the database management saves the patientrelated information and recognition results.Finally,tests were performed on the RGB-D dataset and real people.The results show that the designed and implemented 3D face recognition subsystem based on Kinect has better recognition accuracy and can be applied to the medical insurance review platform.