监狱中人员异常情况视觉监控跟踪方法仿真  

Prison Staff Anomalies Visual Simulation Monitoring Tracking Method

作　　者：杨新锋[1] 刘艳[2] 刘克成[1] 

机构地区：[1]南阳理工学院计算机与信息工程学院,河南南阳473004 [2]中州大学信息工程学院,河南郑州450044

出　　处：《计算机仿真》2015年第7期319-322,共4页Computer Simulation

基　　金：河南省科技攻关计划项目(122102210563)

摘　　要：监狱中人员异常情况视觉监控,可加强对犯罪人员的强制管理。监狱中的人员受到自身条件的限制,犯人会发生异常情况,使得人员运动随机性较大,在监狱中的人员行为特征存在较强的快速突变性,使得对监狱中人员图像特征的提取和跟踪过程必须满足快速变化的需要。传统的视觉监控跟踪方法在运用到监狱中人员异常情况发生的区域中,无法避免由于监狱中人员快速移动造成行为迅速转换,带来的跟踪丢失问题,无法大幅简化跟踪过程,使得经常不能对监狱中人员进行实时准确的视觉监控跟踪。提出采用模糊区域识别跟踪算法的监狱中人员异常情况视觉监控跟踪方法。首先对异常情况发生区域中的监狱中人员图像进行分割,筛选出优质的小特征,降低运算成本,为监狱中人员图像的特征提取提供了准确的数据基础。针对视觉监控图像中部分监狱中人员身体纹理特征进行提取,建立跟踪模型。仿真结果表明,利用改进算法能够有效提高监控的实时性,能满足监狱中人员异常情况视觉监控跟踪的要求。Study of prison staff visual surveillance tracking method,crime of mandatory management personnel has certain significance. Personnel subject to the conditions of their own conditions in prison,the prisoner exception happens,makes the personnel movement randomness,strong fast respectively,is found in the prison for prison staff image features extraction and tracking process must meet the needs of rapid change. Traditional visual surveillance tracking method in applied to prison staff unusual area,inevitable caused by prison staff in fast moving areas lost track of the problems of rapid transformation,not greatly simplify the tracking process,makes for prison staff is often not accurate visual surveillance and tracking in real time. Identification is put forward based on fuzzy area tracking algorithm in prison staff anomalies visual surveillance tracking method. First to unusual prison staff of the area of the image segmentation,select high- quality small features,reduce operation cost,image feature extraction for prison personnel provides accurate data base. In visual surveillance in the image part of the prison staff body texture feature is extracted,tracking model is set up. The simulation results show that the improved algorithm can effectively improve the real-time monitoring,can satisfy the prison staff anomalies in visual surveillance tracking requirements.

关 键 词：监狱 视觉监控 模糊识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]