摘要
针对远程智能监控的应用,基于H.264视频编解码与运动检测标识技术,在由小型工控机与计算机组建的客户机/服务器系统上设计并实现了实时运动检测视频监控系统。系统对采集的视频帧进行H.264规范的压缩编码并传输,在保证图像质量前提下降低了带宽与存储开销。运动检测算法在Kalman动态背景更新的基础上,采用了加权方式的图像差分法。通过实际试验表明,该系统能稳定的压缩传输视频流,并在不增加虚警的情况下,提高了系统的检测准确度。
By using H. 264 video codec and motion detection technology, a realtime motion detection video surveillance system is designed and implemented by Client./Server system built on a tiny industrial machine and a computer. The system compresses the video in H. 264 with higher coding efficiency, maintains the image quality and reduces the requirements of bandwidth and storage. The motion detection algorithm is based on the frame differencing method and Kalman filter based dynamic background updating method. By several actual tests, the system could compress and transmit the video stably and the algorithm could improve the detection accuracy without increasing false detection.
出处
《南京邮电大学学报(自然科学版)》
北大核心
2012年第1期1-5,共5页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
基金
国家科技重大专项(2010ZX03006-004)
国家重点基础研究发展计划(973计划)(2011CB302906)资助项目