• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Unattended ground, sea, and air sensor technologies and applications XII >Time-Domain Classification of Humans using Seismic Sensors
【24h】

Time-Domain Classification of Humans using Seismic Sensors

机译：使用地震传感器的人的时域分类

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

Methods of human classification and direction of travel are developed for the purpose of being embedded in low-power, low-cost microprocessors. Techniques are explored for classifying an impulsive set of events in a seismic field as being either human or non-human based on information extrapolated from time-domain data of geophones. Additionally, a method of time domain direction of travel determination is explored. As a target is traversing the field of detection, simple impulse detection techniques determine seismic activities that are of interest. By recreating the time-domain signal as an average energy over time, the frequency of footstep of the target can be determined after a human has left the field by using post processing techniques, even when multiple targets are present. An autocorrelation of the energy averaged signal will yield an output that can be used to easily determine the most dominant frequency of the observed series of impulsive events. This method is capable of classifying humans under certain conditions at a rate of up to 98% with a varying rate of rejection for different types of animals and environmental factors. The technique can be easily integrated to work in conjunction with other modalities for an increase in classifier confidence.
机译：人们开发了人类分类和行进方向的方法，目的是将其嵌入低功耗，低成本的微处理器中。基于从地震检波器的时域数据推断出的信息，探索了用于将地震场中的脉冲事件集分类为人还是非人的技术。另外，探索了时域行进方向确定的方法。当目标穿越检测领域时，简单的脉冲检测技术将确定感兴趣的地震活动。通过将时域信号重新创建为时间上的平均能量，即使存在多个目标，也可以使用后处理技术确定目标离开人的脚步出现频率。能量平均信号的自相关将产生一个输出，可用于轻松确定所观察到的一系列脉冲事件的最主要频率。该方法能够在某些条件下以高达98％的比率对人类进行分类，并且对不同类型的动物和环境因素的排斥率有所不同。可以轻松集成该技术以与其他模式结合使用，以提高分类器的置信度。

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