Design Of Teaching Assistant System Based On Face Recognition

With the continuous development of our country,in order to meet the needs of my country's rapid development of talents,we need more professional and technical personnel to join the country's construction cause,so as to achieve the expected goals and achieve comprehensive national strength improvement.These professional and technical talents mainly come from our universities,and the way for professional and technical personnel to obtain knowledge is the teaching classroom,so the quality of the teaching classroom directly affects the cultivation of talents.Under the vigorous promotion of the state,the development of college education in my country has made tremendous achievements,accompanied by the expansion of the scale of running schools and changes in the quality of training.Due to the changes in society's demand for talents,colleges and universities are required to cultivate higher quality talents,and the teaching classroom,which is one of the important links of talent training,needs more attention from us.This paper designs a teaching aid system based on face recognition,which is used to record the teaching classroom situation.Through analyzing the data of the teaching classroom situation,we can continuously improve the teaching methods and improve the teaching quality.The research content of this article is as follows:1.Investigate and research the recording method of the existing teaching classroom situation,and design the teaching assistant system based on face recognition.This paper describes the face recognition technology involved,and uses image geometric processing and grayscale processing to create a face sample library.The face features are extracted using Haar features and integral graphs,and the Adaboost algorithm is used to cascade the classifiers to realize the detection of faces in images.The face image is reduced by PCA,and then SVM is used to realize face classification and recognition.In this paper,the involved Adaboost algorithm,PCA algorithm and SVM algorithm are studied respectively,and the results are verified by experimental simulation.2.This system uses the ARM embedded platform as an experimental platform to realize the collection of student face image data.Using the method of installing Ubuntu in a virtual machine,an embedded experimental platform debugging system was built,and a cross-compilation environment and an installation program runtime library were built on the platform.We have carried out version selection,compilation and debugging of the startup program,kernel kernel and root file system required by the experimental platform,and implemented the writing of system file images through OTG to complete the system migration work,which provided hardware conditions for our later experiments.3.The software client implements the functions of student face image acquisition module,classroom situation recording module,and data query module through multithreading technology.The image acquisition module uses the V4L2 framework,and the server transmits the collected image data to the client through the wired network.The network connection is mainly realized by the socket network communication mechanism under the Linux system.In order to facilitate the processing of face images,the OpenCV development environment was built,and under the built debugging environment,the functions of training the student identification model,student identification,recording of the identification situation,and image data preservation were realized.4.Finally,the system software and hardware are tested.The test content includes identity verification,network connection,data transmission,face image collection,identity recognition,data query,etc.The system can achieve teaching through the way of student identity recognition and classroom image data storage Recording function of classroom situation.In addition,we can analyze the situation of students participating in the teaching classroom through the data query function and the saved classroom teaching videos.By testing these functions,the feasibility of the system design is verified.