Off the beaten track: Predicting localisation performance in visual teach and repeat
Off the beaten track: Predicting localisation performance in visual teach and repeat
复制标题
DOI：
10.1109/icra.2016.7487209
复制
发表时间：
2016-05
期刊：
影响因子：
--
通讯作者：
J. Dequaire;Chi Hay Tong;W. Churchill;I. Posner
中科院分区：
文献类型：
--
作者：
J. Dequaire;Chi Hay Tong;W. Churchill;I. Posner
本文提出了一种基于外观的方法来评估视觉教学和重复背景下的定位表现。具体来说，它旨在估计教学轨迹周围的可能走廊，在这个走廊内，基于视觉的定位系统仍然能够定位自己。与现有技术相比，我们的系统能够预测类似轨迹的定位包络，但地理位置遥远，尚未进行重复运行。因此，通过描述一个地区的本地化表现，我们能够预测另一个地区的表现。为了实现这一点，我们利用高斯过程回归器来估计教学运行中任何关键帧的特征匹配的可能数量，基于曲率和关键帧的外观模型等轨迹属性的组合。使用来自真实遍历的数据，我们证明了我们的方法在基于大量重复运行的插值定位性能方面表现得与现有技术一样好，同时在将性能估计泛化到新教授的轨迹方面也表现良好。
This paper proposes an appearance-based approach to estimating localisation performance in the context of visual teach and repeat. Specifically, it aims to estimate the likely corridor around a taught trajectory within which a vision-based localisation system is still able to localise itself. In contrast to prior art, our system is able to predict this localisation envelope for trajectories in similar, yet geographically distant locations where no repeat runs have yet been performed. Thus, by characterising the localisation performance in one region, we are able to predict performance in another. To achieve this, we leverage a Gaussian Process regressor to estimate the likely number of feature matches for any keyframe in the teach run, based on a combination of trajectory properties such as curvature and an appearance model of the keyframe. Using data from real traversals, we demonstrate that our approach performs as well as prior art when it comes to interpolating localisation performance based on a number of repeat runs, while also performing well at generalising performance estimation to freshly taught trajectories.