首页> 外文会议>NATO advanced research workshop on multisensor data fusion >OBJECT DETECTION AND TRACKING IN DISTRIBUTED SURVEILLANCE SYSTEMS USING MULTIPLE CAMERAS Object Detection and Tracking in Distributed Surveillance Systems Using Multiple Cameras
【24h】

OBJECT DETECTION AND TRACKING IN DISTRIBUTED SURVEILLANCE SYSTEMS USING MULTIPLE CAMERAS Object Detection and Tracking in Distributed Surveillance Systems Using Multiple Cameras

机译:使用多个摄像机的分布式监控系统中的多个摄像机对象检测和跟踪分布式监控系统的对象检测和跟踪

获取原文

摘要

In this paper, the main problems related to the design of the architecture of a Distributed Multisensor Surveillance System have been highlighted. Processing and communication aspects have been considered by providing an insight in several issues and related solutions. The paper has stressed from one side the high variability of possible solutions, while highlighting the possibility of organizing in a general framework the complex set of knowledge and expertise necessary in a multidisciplinary field such as surveillance. It has been shown that, in this context, data fusion methods play a very important role, as they are constitute a key step to allow spatial-temporal fusion occurring at the physical level of the architecture as well as fusion related to attributes of events occurring at the logical level. In this direction, examples have been provided describing new emerging methods and related applications, where either multiple cameras or multiple knowledge sources have been employed. Agent-based fusion has been presented as an example of spatial-temporal information fusion, as well as integration of multiple views to form a mosaicing provided by a physical virtual sensor from views coming from a mobile camera. An example describing fusion of shape information associated with an event detected by more cameras (i.e. a person walking in a scene) lias been also presented. A complex system integrating both physical and logical fusion methods in a complex architecture oriented at people number estimation in a wide area scene has been also discussed in the presented framework. The paper is definitively not exhaustive as the global problem is so complex and multidisciplinary that a big amount of research still needs to be carried on: the most interesting issues to be explored in the next years, in the authors' opinion are related to new aspects related to surveillance service providing over open networks, such as internet, as well as to automatic processing tools for ensuring privacy protection issues in DMSSs.
机译:在本文中,突出了与分布式多传感器监控系统的架构设计相关的主要问题。通过在若干问题和相关解决方案中提供洞察,已经考虑了处理和通信方面。本文从一侧强调了可能的解决方案的高度变化,同时突出了一般框架组织的可能性,这些框架组织了多学科领域所需的复杂知识和专业知识,如监视。已经表明,在此上下文中,数据融合方法发挥着非常重要的作用,因为它们是允许在架构的物理级别发生的空间融合以及与发生事件属性相关的融合的空间 - 时间融合的关键步骤在逻辑层面。在此方向上,已经提供了描述新的新出现方法和相关应用程序,其中已经采用了多种摄像机或多个知识源。已经呈现基于代理的融合作为空间 - 时间信息融合的示例,以及多个视图的集成,以形成由来自来自移动相机的视图的物理虚拟传感器提供的镶嵌。描述与由更多摄像机检测到的事件相关联的形状信息的融合的示例(即,在场景中行走的人)LIAS也被呈现。在框架的框架中还讨论了一系列集成在广域场景中的人数估计的复杂架构中的复杂系统已经讨论过。当全球问题如此复杂和多学科仍然需要进行大量的研究:仍然需要进行大量的研究:在作者的意见中探讨的最有趣的问题与新的方面有关与监视服务相关的监督服务,提供互联网等开放网络,以及用于确保DMSS中隐私保护问题的自动处理工具。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号