LiDAR-guided object search and detection in Subterranean Environments

TL;DR

This paper introduces a multi-modal sensor approach combining LiDAR and vision to improve object detection in subterranean environments, verified on robots and datasets from underground rescue scenarios.

Contribution

It presents a novel method that uses sparse LiDAR data to generate object proposals and guides a PTZ camera for directed search in challenging underground conditions.

Findings

Effective object detection at longer distances in subterranean environments

Successful deployment on an ANYmal quadruped robot

Validated on DARPA Subterranean Challenge datasets

Abstract

Detecting objects of interest, such as human survivors, safety equipment, and structure access points, is critical to any search-and-rescue operation. Robots deployed for such time-sensitive efforts rely on their onboard sensors to perform their designated tasks. However, as disaster response operations are predominantly conducted under perceptually degraded conditions, commonly utilized sensors such as visual cameras and LiDARs suffer in terms of performance degradation. In response, this work presents a method that utilizes the complementary nature of vision and depth sensors to leverage multi-modal information to aid object detection at longer distances. In particular, depth and intensity values from sparse LiDAR returns are used to generate proposals for objects present in the environment. These proposals are then utilized by a Pan-Tilt-Zoom (PTZ) camera system to perform a directed search by adjusting its pose and zoom level for performing object detection and classification in difficult environments. The proposed work has been thoroughly verified using an ANYmal quadruped robot in underground settings and on datasets collected during the DARPA Subterranean Challenge finals.