Research on Automatic Algorithm for Counting the Number of People in Classrooms Based on Multi-Object Tracking
一磊 王 · Artificial Intelligence and Robotics Research · 2025
本文聚焦于基于多目标跟踪的课堂人数自动统计这一主题，提出将YOLOv8和SORT (Simple Online and Realtime Tracking)算法相结合的方法。其中，YOLOv8作为先进的单阶段目标检测算法，能迅速识别图像或视频帧中的目标并输出关键信息；SORT算法借助卡尔曼滤波预测目标位置变化。研究通过运用特定的软硬件配置开展实验，对模型进行训练与验证。该方法旨在实现课堂人数的精准自动统计，满足课堂管理和动态监控的需求，为教师及管理人员提供有力工具。其不仅在算法应用上具有创新性，而且在教育管理实践中具有重要的实用价值，有望对智慧教育领域的课堂管理产生积极深远的影响。This paper focuses on the theme of automatic counting of the number of people in the classroom based on multi-object tracking and proposes a method that combines the YOLOv8 and SORT algorithms. Among them, YOLOv8, as an advanced single-stage target detection algorithm, can quickly identify targets in an image or video frame and output key information. The SORT algorithm predicts the position change of the target with the help of Kalman filtering. The research conducts experiment by using specific software and hardware configurations to train and validate the model. This method aims to achieve accurate automatic counting of the number of people in the classroom, meet the needs of classroom management and dynamic monitoring, and provide a powerful tool for teachers and administrators. It is not only innovative in algorithm application, but also has important practical value in educational management practice, and is expected to have a positive and far-reaching impact on classroom management in the field of smart education.