MatchPlant: An Open-Source Pipeline for UAV-Based Single-Plant Detection and Data Extraction

2025-06-14

AI Summary

Accurate single-plant detection in UAV imagery remains challenging due to geometric distortions, scale variation, and labor-intensive annotation. Method: We propose an open-source, modular, GUI-driven pipeline integrating OpenCV-based preprocessing, YOLO/RetinaNet object detection, orthomosaic geometric rectification, IoU-aware geospatial projection, and GDAL-based vectorization. It introduces interactive annotation with cross-temporal detection result reuse and automated plant-height/NDVI extraction. Contribution/Results: Our method achieves seamless, high-accuracy mapping from detection bounding boxes to geographic space (87.5% of projections yield IoU > 0.5) and embeds geospatial analytics. Evaluated on early-stage maize, it attains AP = 89.6% (85.9% on test set), with phenotypic parameters strongly correlated (r = 0.87–0.97) against ground truth and covering 89.8% of manual annotations—substantially reducing annotation effort and cost.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Accurate identification of individual plants from unmanned aerial vehicle (UAV) images is essential for advancing high-throughput phenotyping and supporting data-driven decision-making in plant breeding. This study presents MatchPlant, a modular, graphical user interface-supported, open-source Python pipeline for UAV-based single-plant detection and geospatial trait extraction. MatchPlant enables end-to-end workflows by integrating UAV image processing, user-guided annotation, Convolutional Neural Network model training for object detection, forward projection of bounding boxes onto an orthomosaic, and shapefile generation for spatial phenotypic analysis. In an early-season maize case study, MatchPlant achieved reliable detection performance (validation AP: 89.6%, test AP: 85.9%) and effectively projected bounding boxes, covering 89.8% of manually annotated boxes with 87.5% of projections achieving an Intersection over Union (IoU) greater than 0.5. Trait values extracted from predicted bounding instances showed high agreement with manual annotations (r = 0.87-0.97, IoU>= 0.4). Detection outputs were reused across time points to extract plant height and Normalized Difference Vegetation Index with minimal additional annotation, facilitating efficient temporal phenotyping. By combining modular design, reproducibility, and geospatial precision, MatchPlant offers a scalable framework for UAV-based plant-level analysis with broad applicability in agricultural and environmental monitoring.

Problem

• Accurate single-plant detection from UAV images
• Automated geospatial trait extraction for phenotyping
• Scalable framework for plant-level agricultural monitoring

Innovation

• Open-source Python pipeline for plant detection
• Integrates CNN model training and UAV processing
• Modular design with geospatial trait extraction