基于辅助区域的公共环境中目标协同跟踪研究

项目名称： 基于辅助区域的公共环境中目标协同跟踪研究

项目编号： No.61203272

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 王江涛

作者单位： 淮北师范大学

项目金额： 24万元

中文摘要： 在校园、广场及街道等公共场合空间中，对特定目标进行持续的、长距离的稳健跟踪是目标跟踪系统的主要任务之一，也是当前公共安全和突发事件处理的迫切需要，具有重要的研究意义和价值。针对公共场合往往存在多个与跟踪目标具备相似外观的干扰物体以及目标容易发生遮挡等困难，本研究拟通过设计伪目标区域和支持元素区域两类辅助区域来对目标所处环境内容进分析，借助于两类区域信息的挖掘对跟踪场景下的不确定因素进行可靠把握。进而在研究两类区域的实时搜索和自适应更新技术的基础上，设计一种将辅助区域信息和跟踪目标高效融合的协同跟踪框架，以提高目标跟踪系统在公共空间中的鲁棒性,为跟踪系统的实用化打下坚实的理论基础。

中文关键词： 目标跟踪；区域鉴别分析；粒子滤波；公共环境；

英文摘要： One of the main task for object tracking system is to robustly tracking specific target over long time and long distance in public environment,such as campus,square and streets; this is also the urgent need for public security and handling emergent events, so this research has important significance and value. However, in public environment there usually have several distracters with similar appearance of the real target, and the target also may be subjected to occlusions. To overcome these dilemmas, in this research two type of auxiliary regions: false target regions and supporter regions, are designed to analysis the content of the tracking environment. The uncertain information in the current tracking scenario is reliably learned by means of mining the data of these two type regions. On the basis of developing the real-time searching and adaptive updating techniques for these regions，we tend to provide an cooperative tracking framework which can fusing the information of auxiliary regions and the target efficiently, so as to improve the robustness of object tracking in public environment and lay a good theoretical basis for the tracking system's practicality.

英文关键词： object tracking；region discriminative analysis；particle filter；public environment；