Research On The Application Of Face Recognition In Complex Environments Based On Deep Learning

Now with the rapid development of technology,people's living standards are increasing day by day,and the demand for quality of life is gradually increasing.Therefore,when people pursue the convenience of life,they also have new requirements for safety.Among emerging technologies,face recognition technology meets exactly the requirements of the above two points.At the same time,face recognition as a research focus in the field of pattern recognition and artificial intelligence has always been one of the hot research topics.However,many of today's face recognition systems are aimed at close-range single-person face recognition.These traditional face recognition systems respond to various interference factors such as occlusion,multi-angle side faces,and dark background in complex environments,antiinterference ability is weak.Therefore,it cannot respond to real needs.This paper designs a face recognition system based on deep learning in a variety of interference factors to solve the above problems.In order to ensure the recognition rate of face recognition system under various interference factors,the face recognition system designed in this paper is based on deep learning method.The deep learning framework adopted by the system is deep convolutional neural network Inception-ResNet-v1 for face classification.In face detection and tracking,this paper uses MTCNN network.Image recognition methods such as image sharpening and histogram equalization are used to improve the recognition rate of the classified parts.The system determines the relevant parameters through experiments and tests that its performance is greatly improved compared with the previous convolutional neural network model,and it has better anti-interference ability to interference factors such as occlusion,lighting,and side face changes.The system can realize face recognition under certain complicated conditions.Although using deep learning networks for face recognition can ensure the recognition rate of the system,training with deep neural networks requires large amounts of data and labeled accurate data sets.In actual applications,it may not meet the needs.Therefore,in the design process of the system,it is divided into three parts,namely the data collection part,which is used to collect the data set in the face recognition system;the second part is the model training part,which is used by the face recognition system to collect the good data.The data is trained,and the user data set with a small amount of data can be transferred and learned to solve the problem that the user data is too small to meet the requirements of deep learning.Finally,for the classification and interaction part,the face recognition system realizes the real-time video stream through the human-computer interaction interface Face recognition.This design makes the face recognition system more intuitive and easier to operate.The model is verified through a large number of experiments to determine the parameters of the system to obtain the best parameters.And through the comparison of different face recognition systems,the superiority of the system is verified.In this paper,through the analysis of face recognition technology and development trend,a face recognition system that can be designed under various environmental interference factors is designed.This system meets the practical needs and solves some problems encountered in the practical application of today's face recognition system,and has application value and application prospect.