ROW-SLAM: Under-Canopy Cornfield Semantic SLAM

TL;DR

This paper introduces ROW-SLAM, a semantic SLAM system for autonomous weeding in cornfields, utilizing multi-camera setups and semantic features to improve localization and detection under challenging conditions.

Contribution

It presents a novel multi-camera semantic SLAM approach tailored for under-canopy cornfield environments, addressing challenges of limited space and restricted camera motion.

Findings

Successful field trials across multiple cornfields

Robust detection of corn stalks and ground features

Effective localization in complex agricultural environments

Abstract

We study a semantic SLAM problem faced by a robot tasked with autonomous weeding under the corn canopy. The goal is to detect corn stalks and localize them in a global coordinate frame. This is a challenging setup for existing algorithms because there is very little space between the camera and the plants, and the camera motion is primarily restricted to be along the row. To overcome these challenges, we present a multi-camera system where a side camera (facing the plants) is used for detection whereas front and back cameras are used for motion estimation. Next, we show how semantic features in the environment (corn stalks, ground, and crop planes) can be used to develop a robust semantic SLAM solution and present results from field trials performed throughout the growing season across various cornfields.