A Vision-Based Navigation System for Arable Fields

TL;DR

This work tackles autonomous navigation in arable fields with affordable, vision-based sensing. It develops a complete field-scale pipeline that combines RGB-based in-row navigation with a depth-enabled, vision-based row-switching module, underpinned by a U-Net crop row detector and the Triangle Scan Method, all guided by an IBVS controller. Key contributions include a fully vision-based navigation framework, a vision-driven crop row switching pipeline, and an initial-turning-direction detector, validated through real-field trials and simulated field-scale tests. The results show promising cross-track and heading accuracy over thousands of meters, with GNSS-free switching demonstrated in controlled settings, suggesting practical potential for low-cost autonomous field operations. The work also identifies limitations related to large crop-row gaps and headland irregularities, and proposes sensor fusion and algorithmic enhancements to enable robust, scalable deployment.

Abstract

Vision-based navigation systems in arable fields are an underexplored area in agricultural robot navigation. Vision systems deployed in arable fields face challenges such as fluctuating weed density, varying illumination levels, growth stages and crop row irregularities. Current solutions are often crop-specific and aimed to address limited individual conditions such as illumination or weed density. Moreover, the scarcity of comprehensive datasets hinders the development of generalised machine learning systems for navigating these fields. This paper proposes a suite of deep learning-based perception algorithms using affordable vision sensors for vision-based navigation in arable fields. Initially, a comprehensive dataset that captures the intricacies of multiple crop seasons, various crop types, and a range of field variations was compiled. Next, this study delves into the creation of robust infield perception algorithms capable of accurately detecting crop rows under diverse conditions such as different growth stages, weed density, and varying illumination. Further, it investigates the integration of crop row following with vision-based crop row switching for efficient field-scale navigation. The proposed infield navigation system was tested in commercial arable fields traversing a total distance of 4.5 km with average heading and cross-track errors of 1.24° and 3.32 cm respectively.

Figures (19)

• Figure 1: Camera positioning for Triangle Scan Method. The principal axis of the camera(blue) must always reside within Q2.
• Figure 2: Hexman Mark-1 robot in the Sugar Beet Field.
• Figure 3: Vision-based infield navigation scheme. Green: In-row navigation behaviour, Blue: Row switching behaviour.
• Figure 4: The proposed crop row following architecture with U-Net CNN for crop row mask detection. The crop mask generated by U-Net CNN is used by a triangle scan method to predict a central crop row ($\Delta \theta$: Crop row angle error corresponding to vertical axis, $\Delta L_{x2}$: Positional error of the central crop row relative to image midpoint). de2023deep
• Figure 5: Regions of interest for scanning the top and bottom points of central crop row. Vanishing point scan ROI: RED, Bottom point scan ROI: Green, H: Height of the image, h: Vanishing point scan ROI height. de2023deep