Research on Classroom Behavior Analysis and Quantitative Evaluation System of Student Attention Based on Computer Vision

With the rapid development of computer vision technology, its applications have gradually penetrated into various aspects of daily life. In the field of education, the application of computer vision is still in the early exploration stage. Its application areas mainly focus on experimental platforms, scoring, and attendance systems. In view of the increasingly common smart online education environment, how to apply learner affective computing to the online education environment, in-depth understanding of the learner's learning emotional state is the key issue to improve the quality and effect of online education, and to solve the above problems, the most important means is to use artificial intelligence, big data, computer vision and affective computing and other emerging information technology to track the learner's learning process, accurately analyze and intelligently detect the learner's multimodal emotional information, through identification, fusion, Calculate and visualize learner emotional states, Provide accurate process evaluation of learners' learning effectiveness, provide intelligent feedback and intervention services, and promote students' timely learning reflection to correct learning behaviors and achieve optimal learning results. Using the Analytic Hierarchy Process (AHP), the system integrates multiple evaluation factors with weighted analysis, and applies the PAD model for emotional quantification. The final result provides a numerical standard for evaluating student attention during class. Experimental results demonstrate that the system can accurately assess student engagement, providing effective references for teachers and supporting students' academic development.