摘要
随着行为理解研究的深入,无主包裹的监测分析受到计算机视觉领域研究者的广泛关注。其核心是利用计算机视觉技术从图像序列中检测、跟踪、识别目标并对其行为进行理解与描述。文中结合智能视频监控的应用需求,为了实现无主包裹识别的智能化,将隐马尔科夫模型方法运用在无主包裹识别中,给出了一个基于HMM的无主包裹识别方案,并进行了图像预处理,对HMM参数训练和识别及其识别性能进行了研究,同时对未来的发展趋势作了总结与阐述。
With the thorough of studying behavior understanding, the detection and analysis of unattended packages is receiving increasing attention from computer vision researchers, it aims at attempting to detect, track and identify people, and more generally, to understand human behaviors, from image sequences involving hurnans. With application requirements of intelligent video surveillance, in order to achieve the intelligent unattended packages recognition, implement the HMM （Hidden Markov Models） method in unattended packages recognition,provide a project of unattended packages recognition on HMM, took on the image preprocessing, the training HMM parameters, and identifying from a test sequence, and have been studied in the recognition capability. At the end of this survey, some discussions on future directions in motion analysis are also provided.
出处
《计算机技术与发展》
2009年第3期28-31,共4页
Computer Technology and Development
基金
上海市科技重点计划项目(06dz150003)
关键词
行为理解
无主包裹
隐马尔科夫
behavior understanding
unattended packages
HMM
