The present invention provides a method for detecting video monitoring objects based on Intrusion Detection for an area to be monitored, whether the prohibition of objects, which comprises the following steps: 1: the use of video surveillance cameras capture background reference image data and the non prohibited motion image data, and storage; 2: extraction of non prohibited spatial object quantity, establish non prohibited the object feature model; 3: at a preset sampling frequency real-time acquisition of current image data as historical image data; 4: extraction step 3 historical image data object characteristics and step 2 non background model, characteristics of non prohibited objects, if, return to step 3; otherwise, the non background objects are prohibited objects stored prohibited objects features: 5 step 5; again the current image data acquisition, based on the object feature storage lock step 4 Suppress objects in the current image and alert them. The invention can effectively improve the detection flexibility of the intrusion object, and enlarge the detection range of the intrusion object.
【技术实现步骤摘要】
【技术保护点】
基于视频监控入侵物体的检测方法，采用视频监控摄像头对待监控区域监控，用于检测待监控区域是否出现禁止物体，其特征在于，基于视频监控入侵物体的检测方法包括以下步骤： 步骤１：采用视频监控摄像头采集背景参考图像数据及非禁止物体运动图像数据， 并进行存储； 步骤２：提取非禁止物体空间特征量，建立非禁止物体特征模型； 步骤３：以预设采集频率实时采集当前图像数据作为历史图像数据； 步骤４：提取步骤３中历史图像数据中非背景的物体特征量与步骤２中非禁止物体特征模型进行匹 配，若匹配返回步骤３；若不匹配，则非背景物体为禁止物体，存储禁止物体特征量执行步骤５； 步骤５：再次采集当前图像数据，依据步骤４存储的物体特征量锁定当前图像中禁止物体并报警。
【技术特征摘要】
【专利技术属性】
技术研发人员：林卫慈，倪兴华，许冬青，任晓，韩吉来，应丽华，贺凌云，
申请(专利权)人：上海新联纬讯科技发展有限公司，
类型：发明
国别省市：31[中国|上海]
还没有人留言评论。发表了对其他浏览者有用的留言会获得科技券。