基于图像处理的人员异常行为监测设计

摘要

The ability of intelligent recognition of video monitoring and accident is improved through design of effective monitoring algorithm for the personnel abnormal behavior. The traditional personnel abnormal behavior monitoring algorithm with video step tracking learning method result in poor visual recognition effect due to the randomness and uncertainty of the staff. A personnel abnormal behavior monitoring algorithm based on video frame image array pixel detection is proposed according to the theory of image processing. A data acquisition model of personnel abnormal behavior characteristics was designed and realized, which is based on the double camera video surveillance. The video frame image pixel arrays is used to extract abnormal behavior features for getting the pixels between neighborhood gray value vectors as the mean structure similarity clustering center. The im⁃provement of personnel abnormal behavior monitoring algorithm was implemented by calculating the estimated values of weighted average positions of all pixel points in personnel abnormal behavior detection image. The simulation results show that the algo⁃rithm for monitoring the characteristics of personnel abnormal behavior can realize the accurate positioning and clear identifica⁃tion of the specific abnormal personnel,and has high detection probability and superior performance. It has high application va⁃lue in the field of security monitoring system design.%通过对人员异常行为有效的监测算法设计，提高视频监控和异常事故的智能识别能力。传统的人员异常行为监测算法采用视频步进跟踪学习方法，由于人员的随机性和不确定性，导致视觉识别效果不好。基于图像处理理论，提出一种基于视频帧图像阵列像素检测的人员异常行为监测算法，进行了人员异常行为特征数据采集模型设计，得到基于双相机视频监控的人员异常行为视觉特征采集模型，采用视频帧图像阵列像素检测算法，进行异常行为特征提取，得到邻域灰度值向量之间像素点为平均结构相似性聚类中心，计算人员异常行为检测全图所有像素点的加权平均位置的估计值，实现人员异常行为监测算法改进。仿真结果表明，采用该算法进行人员异常行为特征监测，能实现对特定异常人员的准确定位和清晰识别，异常检测概率较高，性能优越，在安防监控系统设计等领域具有较好的应用价值。