考场实时智能巡查系统的构建与应用  

The Development and Implementation of Real-Time Intelligent Inspection System for Examination Venue

作　　者：赵康 王巍 丛涌泉 刘洪通 朱坤昂 ZHAO Kang;WANG Wei;CONG Yongquan;LIU Hongtong;ZHU Kunang(Shandong Provincial Academy of Educational Recruitment and Examination,Jinan 250011,China;Qingdao Municipal Academy of Recruitment and Examination,Qingdao 266071,China)

机构地区：[1]山东省教育招生考试院,济南250011 [2]青岛市招生考试院,山东青岛266071

出　　处：《中国考试》2025年第4期51-59,共9页journal of China Examinations

摘　　要：考场行为管理是考试安全的重要保障。本研究应用实时目标检测、行为识别等人工智能技术,基于考场视频采集与分析并融合考务规范与人机协同,构建了一套全流程考场实时智能巡查体系,为监考工作提供过程性决策支撑。通过高考大规模现场应用实践,验证了智能行为识别技术在考场行为管理中实际应用的可行性,以及其在警示考生、规范监考、提高效率、确保安全等方面的积极作用。基于面向考场行为规范分析判定的应用实效研究,可以推动针对考场实时智能巡查技术的持续改进,为人工智能进一步助力考试管理、保障考试安全积累经验。Examination venue behavior management serves as a critical safeguard for examination security.This study develops an end-to-end intelligent inspection system for examination venues by leveraging cutting-edge artificial intelligence technologies,including real-time object detection and multimodal behavior recognition.The system integrates video data collection and analysis with examination regulations and human-machine collaboration,providing procedural decision-making support for invigilation tasks,effectively enhancing the quality of examination management and mitigating security risks.The feasibility of real-time behavior recognition technology in practical examination venue management has been validated through large-scale field applications during the Gaokao.Furthermore,the system has proven effective in warning candidates,standardizing invigilation protocols,enhancing operational efficiency,and ensuring safety.By assessing the practical effectiveness of behavioral norm assessment within examination environments,this research drives continuous advancements in real-time intelligent inspection technology and provides valuable insights for further integration of artificial intelligence into examination administration and security assurance.

关 键 词：考场监控 高考 人工智能 行为识别 智能巡查 

分 类 号：G405[文化科学—教育学原理]