Development of QR Code-Guided Autonomous Navigation System for Grape Cultivation Robot in Overhead Trellis Vineyard

This paper presents a novel autonomous navigation system for grapevine cultivation robots operating in overhead trellis vineyards, a challenging environment characterized by irregular grape cluster distribution and complex 3D structures. The system integrates two complementary navigation approaches: high-precision positioning under target grape bunches using YOLOv8-based object detection of QR codes, and large-scale vineyard navigation using global navigation satellite system (GNSS). Each QR code uniquely identifies a grape bunch and provides access to its associated cultivation data. Experimental evaluations in a simulated vineyard environment demonstrated an 83.3% success rate for object detection-based positioning beneath grape bunches and 90% accuracy for GNSS-based navigation within a 1.0 m radius of the target locations. These results validate the system's effectiveness in automating navigation tasks within overhead trellis vineyards, offering a promising solution to address labor shortages and improve the efficiency of grape cultivation practices.