• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Definition of algorithms for characterization and matching magnetic fingerprints to re-identify vehicles at interurban roads based on magneto-resistive sensors.
【24h】

Definition of algorithms for characterization and matching magnetic fingerprints to re-identify vehicles at interurban roads based on magneto-resistive sensors.

机译：定义用于表征和匹配磁性指纹的算法，以基于磁阻传感器在城市间道路上重新识别车辆。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Traffic Management is a task that needs as much information as possible about the roads status and about the vehicles that are circulating on them. This information is usually obtained by processing the data that can be extracted from a large set of sensors located around the roads and from the circulating vehicles. In this context, the most appreciated data is that about the individual paths of each and every vehicle. However, few are the types of sensors that can provide the traffic managers with this kind of data by means of a re-identification process at different points of the road network.;Magnetic sensors have been applied to traffic detection since the 1960's. The evolution of the technology in the most recent years has brought a new type of sensors that can extract very detailed information from the moving vehicles: the magneto-resistive sensors. These sensors can obtain a particular characterization of every vehicle through a magnetic fingerprint. The magnetic fingerprint is just the representation of the interference that a moving vehicle causes to the Earth magnetic field.;Magnetic fingerprints from magneto-resistive sensors have been recently used for vehicle classification and re-identification purposes. The work of scientists in this field shows the difficulties of getting accurate data for these objectives. The major problems that have been found are the acquisition of magnetic fingerprints for all vehicles (detection), the extraction of relevant information from the magnetic time series (signal segmentation), the selection of the appropriate data of each signal for a particular problem (signal processing), the use of an optimal method to compare vehicles through their magnetic fingerprints (similarity measure), the definition of a sensors architecture to gather the relevant information from the vehicles (sensor network) and the consideration of a magnetic model for the vehicles during the re-identification process (magnetic characterization of the vehicles).;This thesis deals with the above mentioned problems and proposes new solutions to improve the vehicles re-identification ratios for inter-urban roads. In this environment, the high vehicles speeds and varying trajectories increase the problem complexity in contrast to the urban roads, where almost all the previous studies have been carried out.;Firstly, a study of the signals produced by the magneto-resistive sensors will be made, together with an analysis of the similarity measures that have been used by other researchers with re-identification purposes. The result of the study will be a new signal extraction and comparison method, the definition of a procedure for an adequate signal processing phase, and the justified definition of a similarity measure to compare vehicles. A signal alignment technique will also be proposed in order to improve the similarity values and the vehicles re-identification ratios.;Secondly, a vehicle magnetic model will be studied by using the data gathered by the magneto-resistive sensors. The complexity of the vehicles structure will be analyzed and the longitudinal vehicle section that can produce the best signals for re-identification will be bounded.;Finally, an experiment with real traffic conditions will be made (M-12 highway). A new configuration for a sensors network will be tested. The algorithms and procedures defined during the previous work and experiments will be put into practice and tested. The results of the final experiment will show the improvements that the proposed methodology has on vehicle re-identification performance, compared to previous research works.
机译：交通管理是一项任务，需要尽可能多的信息，包括道路状况以及在道路上行驶的车辆。通常，通过处理可从位于道路周围的大量传感器以及正在行驶的车辆中提取的数据来获取此信息。在这种情况下，最受赞赏的数据是关于每辆车的各个路径的数据。但是，很少有传感器能够通过重新识别道路网络的不同点来为交通管理者提供此类数据的传感器。自1960年代以来，磁性传感器已应用于交通检测。近年来，技术的发展带来了一种新型传感器，它可以从行驶中的车辆中提取非常详细的信息：磁阻传感器。这些传感器可以通过磁性指纹获得每个车辆的特定特征。磁性指纹只是行驶中的车辆对地磁场造成的干扰的表示。磁阻传感器的磁性指纹最近已用于车辆分类和重新识别。在这一领域的科学家的工作表明，很难获得用于这些目标的准确数据。已发现的主要问题是：获取所有车辆的磁指纹（检测），从磁时间序列中提取相关信息（信号分割），针对特定问题（信号）选择每个信号的适当数据处理），使用最佳方法通过其磁性指纹比较车辆（相似性度量），定义传感器架构以从车辆收集相关信息（传感器网络）以及在行驶过程中考虑车辆的磁性模型本文针对上述问题，提出了新的解决方案，以提高城市间道路车辆的重新识别率。在这种环境下，与以前几乎所有研究都在进行的城市道路相比，高车速和变化的轨迹增加了问题的复杂性。首先，将对磁阻传感器产生的信号进行研究。并分析了其他研究人员用于重新识别目的的相似性度量。研究的结果将是一种新的信号提取和比较方法，适当信号处理阶段程序的定义，以及用于比较车辆的相似性度量的合理定义。还将提出一种信号对准技术，以提高相似度值和车辆的重新识别率。其次，将利用磁阻传感器收集的数据研究车辆的磁模型。将分析车辆结构的复杂性，并限制能够产生最佳信号以进行重新识别的纵向车辆部分。最后，将进行具有实际交通状况的实验（M-12高速公路）。将测试传感器网络的新配置。在先前的工作和实验中定义的算法和过程将付诸实践并进行测试。与之前的研究工作相比，最终实验的结果将表明所提出的方法在车辆重新识别性能方面的改进。

著录项

  • 作者

    Gimeno, Ramon Vicente Cirilo.;

  • 作者单位

    Universitat de Valencia (Spain).;

  • 授予单位 Universitat de Valencia (Spain).;
  • 学科 Computer engineering.
  • 学位 Dr.
  • 年度 2015
  • 页码 431 p.
  • 总页数 431
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

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