Formation of a Lightweight, Deep Learning-Based Weed Detection System for a Commercial Autonomous Laser Weeding Robot

Abstract

Weed management is becoming increasingly important for sustainable crop production. Weeds cause an average yield loss of 11.5% billion in Pakistan, which is more than PKR 65 billion per year. A real-time laser weeding robot can increase the crop’s yield by efficiently removing weeds. Therefore, it helps decrease the environmental risks associated with traditional weed management approaches. However, to work efficiently and accurately, the weeding robot must have a robust weed detection mechanism to avoid physical damage to the targeted crops. This work focuses on developing a lightweight weed detection mechanism to assist laser weeding robots. The weed images were collected from six different agriculture farms in Pakistan. The dataset consisted of 9000 images of three crops: okra, bitter gourd, sponge gourd, and four weed species (horseweed, herb paris, grasses, and small weeds). We chose a single-shot object detection model, YOLO5. The selected model achieved a mAP of 0.88@IOU 0.5, indicating that the model predicted a large number of true positive (TP) with much less prediction of false positive (FP) and false negative (FN). While SSD-ResNet50 achieved a mAP of 0.53@IOU 0.5, the model predicted fewer TP with significant outcomes as FP or FN. The superior performance of the YOLOv5 model made it suitable for detecting and classifying weeds and crops within fields. Furthermore, the model was ported to an Nvidia Xavier AGX standalone device to make it a high-performance and low-power computation detection system. The model achieved an FPS rate of 27. Therefore, it is highly compatible with the laser weeding robot, which takes approximately 22.04 h at a velocity of 0.25 feet per second to remove weeds from a one-acre plot.

Model Development for Identifying Aromatic Herbs Using Object Detection Algorithm

“…Thus, there is an opportunity to improve the database for the bay leaf herb and redo the bounding boxes. The challenges previously raised [8,18,34] were noted in the method stage, as image recognition became challenging due to variations in the aromatic herb's size, angle, and quality. These proportions or new or unusual configurations characteristic of aromatic herbs represent a challenge regarding the volume and diversity of the database to be used for training.…”

“…There has been a theoretical basis for recognizing the rhizome and main root in plants through "internal content difference" and "external morphological difference" [17]. Weed detection was previously performed to assist laser weed removal robots [18]. Previous studies [3,19] determined YOLO's adaptability and efficiency in plant and herb identification through image analysis, leveraging its speed and accuracy in object detection tasks.…”

Development of a Detection System for Types of Weeds in Maize (Zea mays L.) under Greenhouse Conditions Using the YOLOv5 v7.0 Model

“…YOLO algorithms are implemented on different platforms for weed identification, as in [22], where unmanned ground vehicles (UGVs) were used for weed identification and removal in lettuce crops using the YOLO, Faster R-CNN, and SSD Mobile models. In [23], he used YOLO-v5 to build a real-time laser weeding robot in three crops: okra, bitter gourd, and sponge squash and four weed species, achieving a mAP of 88%. It has also been used in applications with UAVs (unmanned aerial vehicles); for example, [24] used a UAV to capture images and process them with YOLO-v7 in order to detect the weed Mercurialis annua in a sugar beet crop, obtaining a mAP of 62.1%.…”

Weed detection in soybean fields using improved YOLOv7 and evaluating herbicide reduction efficacy

“…(2022) detected weeds in alfalfa through YOLOv3. Fatima et al. (2023) developed a lightweight weed detection mechanism to assist laser-weeding robots through YOLOv5.…”

“…Other scholars such as Yang et al (2022) detected weeds in alfalfa through YOLOv3. Fatima et al (2023) developed a lightweight weed detection mechanism to assist laser-weeding robots through YOLOv5. Liu M. et al (2022) constructed a weed detection model for maize fields based on the MSRCR-YOLOv4-tiny, which provides a feasible realtime weed identification method for precision weed control systems in fields with limited hardware resources.…”