Intrarow Uncut Weed Detection Using You-Only-Look-Once Instance Segmentation for Orchard Plantations

Rizky Mulya Sampurno,Zifu Liu,R. M. Rasika D. Abeyrathna,Tofael Ahamed

DOI: https://doi.org/10.3390/s24030893

IF: 3.9

2024-01-31

Sensors

Abstract:Mechanical weed management is a drudging task that requires manpower and has risks when conducted within rows of orchards. However, intrarow weeding must still be conducted by manual labor due to the restricted movements of riding mowers within the rows of orchards due to their confined structures with nets and poles. However, autonomous robotic weeders still face challenges identifying uncut weeds due to the obstruction of Global Navigation Satellite System (GNSS) signals caused by poles and tree canopies. A properly designed intelligent vision system would have the potential to achieve the desired outcome by utilizing an autonomous weeder to perform operations in uncut sections. Therefore, the objective of this study is to develop a vision module using a custom-trained dataset on YOLO instance segmentation algorithms to support autonomous robotic weeders in recognizing uncut weeds and obstacles (i.e., fruit tree trunks, fixed poles) within rows. The training dataset was acquired from a pear orchard located at the Tsukuba Plant Innovation Research Center (T-PIRC) at the University of Tsukuba, Japan. In total, 5000 images were preprocessed and labeled for training and testing using YOLO models. Four versions of edge-device-dedicated YOLO instance segmentation were utilized in this research—YOLOv5n-seg, YOLOv5s-seg, YOLOv8n-seg, and YOLOv8s-seg—for real-time application with an autonomous weeder. A comparison study was conducted to evaluate all YOLO models in terms of detection accuracy, model complexity, and inference speed. The smaller YOLOv5-based and YOLOv8-based models were found to be more efficient than the larger models, and YOLOv8n-seg was selected as the vision module for the autonomous weeder. In the evaluation process, YOLOv8n-seg had better segmentation accuracy than YOLOv5n-seg, while the latter had the fastest inference time. The performance of YOLOv8n-seg was also acceptable when it was deployed on a resource-constrained device that is appropriate for robotic weeders. The results indicated that the proposed deep learning-based detection accuracy and inference speed can be used for object recognition via edge devices for robotic operation during intrarow weeding operations in orchards.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The paper aims to solve the detection problem of weeds not removed intrarow in the orchard, especially the interruption of Global Navigation Satellite System (GNSS) signals caused by obstacles such as trees, nets, and poles in the orchard planting structure. Due to the inefficiency of mechanical weeders in tree rows, manual weeding is still required in these areas. The paper proposes a visual module based on the You-Only-Look-Once (YOLO) instance segmentation algorithm, which is trained on a customized dataset to support the automatic driving of weed-removing robots in identifying uncultivated weeds and obstacles (such as tree trunks and fixed poles).

The research team collected 5000 images from a pear orchard at Tsukuba Plant Innovation Research Center (T-PIRC) in Tsukuba, Japan, and performed preprocessing and annotation for training and testing the YOLO model. They compared four YOLO instance segmentation models optimized for edge devices, including YOLOv5n-seg, YOLOv5s-seg, YOLOv8n-seg, and YOLOv8s-seg. The results showed that the smaller YOLOv5 and YOLOv8 models were more efficient in terms of detection accuracy, model complexity, and inference speed, with YOLOv8n-seg being selected as the visual module for the automatic driving weed-removing robot.

With this study, the authors demonstrated that deep learning-based detection accuracy and inference speed can be used for object recognition on edge devices to support robotic weeding operations in orchards, especially in areas where weeds in tree rows are not removed.