Analysis And Evaluation Of Invigilator Behavior Based On Motion Trajectory In The Examination Roo

The analysis of invigilator behavior is significant part of the research in intelligent examination rooms.According to the files in the"adult college entrance examination work regulation in 2015",this paper analyzes and research on the invigilators’ motion trajectory and judge the behavior of the invigilators.The research content involves moving object detection and tracking in complex scenes and fuzzy evaluation.The behavior of intelligent analysis and evaluation of examination invigilation system is designed and Implemented.The system accomplishes the trajectory tracking of the invigilators.Based on the fuzzy evaluation and the tracking trajectory the comprehensive evaluation is obtained.The main research work are as follows.(1)Moving target detection in the examination room.Taking into account that the examination room is relatively complex environment.The subtle movement of the candidates will affect the extraction of the invigilators’ region.This paper combines the Three-Frame-Difference method and ViBe algorithm.On the one hand,Three-Frame-Difference method is not sensitive to the prospect of change.So it can filter out the interference caused by the subtle movements of the candidates.On the other hand,the ViBe algorithm is more accurate in the description of the details in the foreground region.Combining the two characteristics can make the proctors detection more accurate.In addition,the Three-Frame-Difference method is mended to improve its operation speed.It provides a clear target for the subsequent extraction of the target and also ensures the real-time performance.(2)Moving object tracking for invigilating teachers in examination rooms.According to the result of moving target detection,executing feature extraction on invigilating teachers.The two extracting features will be fused.There are two features,one is color correlation histogram feature which includes spatial information,another one is scale invariant feature which is robust to affine transformation.The two features mentioned above will be fused based on fuzzy mathematics.Finally,the fused feature will be applied to particle filter algorithm to complete target tracking.(3)The trajectory tracking and evaluation of invigilating teachers.With the process of analysing the data of trajectory,some unusual data will be fixed.The result of the analysis will service for the evaluation.Classifying the evaluation factors to achieve classification evaluation.By means of analytic hierarchy process,the weight of each evaluation factor is obtained.Finally make a comprehensive evaluation.(4)The designation and implementation of the invigilating behavior analysis and evaluation system.Introducing the environment of experiment operation and related tools and the system development framework.At the same time,the composition of the function modules in the system is introduced.The effectiveness of the system operation in the bitter end.