Positioning radiata pine branches requiring pruning by drone stereo vision

2026-04-12

AI Summary

This study addresses the need for autonomous pruning in radiata pine forestry by proposing an automated branch detection and localization method based on drone-mounted stereo vision. The approach integrates instance segmentation models (YOLOv8/v9 and Mask R-CNN) with depth estimation techniques—including both traditional stereo matching (SGBM) and deep learning-based methods (PSMNet, RAFT-Stereo)—within a two-stage pipeline to achieve pixel-level branch segmentation and 3D spatial localization. A novel centroid triangulation algorithm leveraging segmentation masks and disparity maps is introduced, augmented by the Median Absolute Deviation (MAD) criterion to effectively reject outliers, thereby significantly improving localization accuracy at close ranges (1–2 meters). Experimental results demonstrate that disparity maps generated by deep learning models exhibit superior consistency over short distances compared to conventional approaches, confirming the feasibility of low-cost stereo vision systems for autonomous forestry pruning applications.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Ensemble Methods

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This paper presents a stereo-vision-based system mounted on a drone for detecting and localising radiata pine branches to support autonomous pruning. The proposed pipeline comprises two stages: branch segmentation and depth estimation. For segmentation, YOLOv8, YOLOv9, and Mask R-CNN variants are compared on a custom dataset of 71 stereo image pairs captured with a ZED Mini camera. For depth estimation, both a traditional method (SGBM with WLS filtering) and deep-learning-based methods (PSMNet, ACVNet, GWCNet, MobileStereoNet, RAFT-Stereo, and NeRF-Supervised Deep Stereo) are evaluated. A centroid-based triangulation algorithm with MAD outlier rejection is proposed to compute branch distance from the segmentation mask and disparity map. Qualitative evaluation at distances of 1-2 m indicates that the deep learning-based disparity maps produce more coherent depth estimates than SGBM, demonstrating the feasibility of low-cost stereo vision for automated branch positioning in forestry.

Problem

• radiata pine
• branch positioning
• stereo vision
• autonomous pruning
• drone

Innovation

• stereo vision
• autonomous pruning
• branch segmentation
• depth estimation
• drone-based forestry