原文传递
Image-Based Vehicle Identification Technology for Homeland Security Applications
| 题名： | Image-Based Vehicle Identification Technology for Homeland Security Applications |
| 作者： | Clark, G A; |
| 关键词： | 42 ENGINEERING; LAWRENCE LIVERMORE NATIONAL LABORATORY; MONITORING; PERFORMANCE; SECURITY; TRAINING; URBAN AREAS; VELOCITY; WEAPONS |
| 摘要： | The threat of terrorist attacks against US civilian populations is a very real, near-term problem that must be addressed, especially in response to possible use of Weapons of Mass Destruction. Several programs are now being funded by the US Government to put into place means by which the effects of a terrorist attack could be averted or limited through the use of sensors and monitoring technology. Specialized systems that detect certain threat materials, while effective within certain performance limits, cannot generally be used efficiently to track a mobile threat such as a vehicle over a large urban area. The key elements of an effective system are an image feature-based vehicle identification technique and a networked sensor system. We have briefly examined current uses of image and feature recognition techniques to the urban tracking problem and set forth the outlines of a proposal for application of LLNL technologies to this critical problem. The primary contributions of the proposed work lie in filling important needs not addressed by the current program: (1) The ability to create vehicle ''fingerprints,'' or feature information from images to allow automatic identification of vehicles. Currently, the analysis task is done entirely by humans. The goal is to aid the analyst by reducing the amount of data he/she must analyze and reduce errors caused by inattention or lack of training. This capability has broad application to problems associated with extraction of useful features from large data sets. (2) Improvements in the effectiveness of LLNL's WATS (Wide Area Tracking System) by providing it accurate threat vehicle location and velocity. Model predictability is likely to be enhanced by use of more information related to different data sets. We believe that the LLNL can accomplish the proposed tasks and enhance the effectiveness of the system now under development. |
| 报告类型： | 科技报告 |
相关文献
- Thermal Image-Based Deer Detection to Reduce Accidents Due to Deer-Vehicle Collisions.
- Thermal Image-Based Deer Detection to Reduce Accidents Due to Deer-Vehicle Collisions.
- Image-based ranging and guidance for rotorcraft
- Exploring Image-Based Classification to Detect Vehicle Make and Model. Final Report.
- Department of Homeland Security, Science and Technology Directorate, Borders and Maritime Security Division: Renewable Energy Research Roadmap.
- Modeling Terrorism Risk to the Air Transportation System: An Independent Assessment of TSA's Risk Management Analysis Tool and Associated Methods.
检索历史
应用推荐
