Real-time LIDAR localization in natural and urban environments

Georgi Tinchev, Adrian Penate-Sanchez, Maurice Fallon

[arXiv] [Code (TBA)] [Dataset (TBA)] [Slides (TBA)]

Abstract: Localization is a key challenge in many robotics applications. In this work we explore LIDAR-based global localization in both urban and natural environments and develop a method suitable for online application. Our approach leverages efficient deep learning architecture capable of learning compact point cloud descriptors directly from 3D data. The method uses an efficient feature space representation of a set of segmented point clouds to match between the current scene and the prior map. We show that down-sampling in the inner layers of the network can significantly reduce computation time without sacrificing performance. We present substantial evaluation of LIDAR-based global localization methods on nine scenarios from six datasets varying between urban, park, forest, and industrial environments. Part of which includes post-processed data from 30 sequences of the Oxford RobotCar dataset, which we make publicly available. Our experiments demonstrate a factor of three reduction of computation, 70\% lower memory consumption with marginal loss in localization frequency. The proposed method allows the full pipeline to run on robots with limited computation payload such as drones, quadrupeds, and UGVs as it does not require a GPU at run time.

Algorithm Preview

 

Supplementary Video 1: Method Overview + KITTI, park, forest, industrial trials

Supplementary Video 2: Real time execution with ROS on the Newer College Dataset

Code

We have approval for our university to open source the project and will publish code here when it's tidied up and ready.

Robotcar Dataset Preview

Download (uploading data, check back soon)

Related Article

• Global LIDAR Localization ( older work )