• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>AIAA guidance, navigation, and control conference;AIAA SciTech forum >Enhancing Inertial Navigation with Structure from Motion Trajectory Estimates
【24h】

Enhancing Inertial Navigation with Structure from Motion Trajectory Estimates

机译：通过运动轨迹估计增强结构的惯性导航

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The ubiquitous nature of global positioning systems (GPS) has fostered the widespread integration of navigation into a multitude of applications, both civilian and military. One alternative for ensuring continued flight operations in the absence of GPS is vision-aided navigation, an approach that combines visual cues from a camera with an inertial measurement unit to estimate the navigation states of a moving body. The experiments presented in this paper are based on the Minor Area Motion Imagery dataset that recorded GPS data, inertial measurement data, and electrooptical imagery during flights over Wright-Patterson Air Force Base. The approach applied here combines traditional inertial measurements with trajectory estimates generated by the structure from motion (SfM) algorithm. A characterization of the trajectory estimates made by SfM in this work demonstrates that it provides zero mean velocity errors with a covariance of 20 m/s. The experiments presented here combine the SfM velocity estimates with simulated inertial data for 250 flights and show that this approach can limit the average navigation error to 100 meters over 30 minute trials without GPS, a significant increase over using an IMU alone. The impact of such results highlights the importance of developing robust navigation solutions that integrate modern computer vision approaches as a means of mitigating our increasing reliance on GPS.
机译：全球定位系统（GPS）的无处不在的性质促进了导航到众多申请的广泛整合，包括民用和军事。确保在没有GPS的情况下确保持续飞行操作的替代方案是视觉辅助导航，一种方法将来自具有惯性测量单元的相机与相机中的视觉提示组合以估计移动体的导航状态。本文中提出的实验基于录制GPS数据，惯性测量数据和电光图像的次要区域运动图像数据集，在赖特 - 帕特森空军基地飞行期间。这里应用的方法将传统的惯性测量与来自运动（SFM）算法的结构产生的轨迹估计相结合。在本工作中SFM制造的轨迹估计的表征表明它提供了零平均速度误差，其协方差为20米/秒。此处提出的实验将SFM速度估计与模拟惯性数据相结合了250个飞行，并表明这种方法可以将平均导航误差限制在30分钟内没有GPS的试验中，仅使用IMU的显着增加。这种结果的影响强调了开发强大的导航解决方案的重要性，将现代计算机视觉方法整合为减轻我们越来越依赖GPS的手段。

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