Fig 6 - uploaded by Yuzhen Lu
Content may be subject to copyright.
Examples of weed images with predicted bounding boxes. The bounding boxes with the same color represent the same weed class. The green, red, purple, and orange bounding boxes represent the detection of Purslane, Waterhemp, Spotted Spurge, and Carpetweed, respectively.
Source publication
![Fig. 5. Training curves of mAP@0.5 and mAP@[0.5:0.95] for the top 10...](profile/Yuzhen-Lu-2/publication/367052549/figure/fig4/AS:11431281122223040@1677231972003/Training-curves-of-mAP05-and-mAP05095-for-the-top-10-YOLO-models-for-weed_Q320.jpg)
Weeds are among the major threats to cotton production. Overreliance on herbicides for weed control has accelerated the evolution of herbicide-resistance in weeds and caused increasing concerns about environments, food safety and human health. Machine vision systems for automated/robotic weeding have received growing interest towards the realizatio...
Context in source publication
Context 1
... terms of mAP@[0.5:0.95], except YOLOv3-tiny that gave 68.18 % accuracy, all the models achieved the accuracy ranging from 83.64 % to 89.72 %, with 22 out of 25 achieving the accuracy exceeding 86 %. The lower mAP@[0.5:0.95] values than mAP@0.5 is because the former used higher IoU thresholds (implying more stringent criteria) for AP calculation. Fig. 6 shows the examples of images predicted by YOLOv3-tiny and YOLOv4. Both models yielded visually good predictions for images with diverse and/or cluttered backgrounds, even with densely populated weeds. The predictions on the test images were also compiled into videos for illustration, which are accessible on the GitHub web- site. ...
Citations
... este fue seleccionado debido a que en el trabajo desarrollado por [12] el modelo yolov5s.pt mostro un mejor rendimiento sobre los demás modelos de la familia Yolov5 en el mismo cultivo, adicionalmente en el trabajo de [15] afirma el mismo resultado. ...
Corn (Zea mays L.) is one of the most important cereals worldwide, to maintain crop productivity it is important to eliminate weeds that compete for nutrients and other resources, the eradication of these causes environmental problems through the use of agrochemicals, so the
implementation of technology to mitigate this impact is a challenge. In this work, an artificial vision system was implemented based on the YOLOv5s (You Only Look Once) model, which uses a single convolutional neural network (CNN) that allows differentiating corn from 4 types of weeds, for which a mobile support structure was built to capture images. The performance of the trained model had a value of mAP@05 (mean Average Precision) at a threshold of 0.5 of 83.6%. A prediction accuracy of 97% and a mAP@05 of 97.5% were obtained for the maize class. For the weed classes, Lolium Perenne, Sonchus Oleraceus, Saolanun Nigrum, and Poa Annua obtained an accuracy of 86%, 90%, 78%, and 74%, and a mAP@05 of 81.5%, 90.2%, 76.6% and 72.0%, respectively. The results are encouraging for the construction of a precision weeding system.
... The culmination of this process is the implementation of these trained models in downstream agricultural applications, such as plant health monitoring (Xu et al., 2022a), crop and weed management (Dang et al., 2023;Rai et al., 2023), fruit picking (Chu et al., 2021), and precision livestock farming (Yang et al., 2023c). For instance, in relation to plant health, the models can be utilized to identify early indicators of disease or pest infestations, thus facilitating prompt interventions to reduce crop loss. ...
The past decade has witnessed the rapid development of ML and DL methodologies in agricultural systems, showcased by great successes in variety of agricultural applications. However, these conventional ML/DL models have certain limitations: They heavily rely on large, costly-to-acquire labeled datasets for training, require specialized expertise for development and maintenance, and are mostly tailored for specific tasks, thus lacking generalizability. Recently, foundation models have demonstrated remarkable successes in language and vision tasks across various domains. These models are trained on a vast amount of data from multiple domains and modalities. Once trained, they can accomplish versatile tasks with just minor fine-tuning and minimal task-specific labeled data. Despite their proven effectiveness and huge potential, there has been little exploration of applying FMs to agriculture fields. Therefore, this study aims to explore the potential of FMs in the field of smart agriculture. In particular, we present conceptual tools and technical background to facilitate the understanding of the problem space and uncover new research directions in this field. To this end, we first review recent FMs in the general computer science domain and categorize them into four categories: language FMs, vision FMs, multimodal FMs, and reinforcement learning FMs. Subsequently, we outline the process of developing agriculture FMs and discuss their potential applications in smart agriculture. We also discuss the unique challenges associated with developing AFMs, including model training, validation, and deployment. Through this study, we contribute to the advancement of AI in agriculture by introducing AFMs as a promising paradigm that can significantly mitigate the reliance on extensive labeled datasets and enhance the efficiency, effectiveness, and generalization of agricultural AI systems.
... Algorithm training All models were trained using the default, recommended hyperparameters as specified in the respective repositories (repository links are provided in Table 2). All versions of YOLO tested here included forms of default, in-built augmentation, which affects performance of the model (Dang et al., 2023). Setting hyperparameters for the training of neural network models is a lengthy process and is highly specific to the dataset on which the models are being tested. ...
... In the highly variable, realistic field conditions represented by the collected dataset (variable lighting conditions, soil backgrounds, and crop-weed growth stages), v5-X was capable of recalling up to 81.42% of A. palmeri plants with a precision of 87.15% and mAP@[0.5:0.95] of 66.76% for single class detection. These results are lower than that of Dang et al. (2023), who achieved an mAP@[0.5:0.95] of up to 88.70% with the same v5-X architecture on a 12-class dataset of different weeds. The images in Dang Figure 8 Comparison of class-wise performance for manually grouped (orange) and size grouped eight-class (dark grey) YOLOv5 X models trained at 1280 × 896 image resolution. ...
... The performance on a dataset of diverse growth stages is promising; however, the substantial decline in performance for eight-class models highlights the challenge of the task. The mAP0.5 for v5-X dropped to 58.60%, while the peak performance for eight classes achieved by variant v7 of 61.14%, well below the results achieved by both Barnhart et al. (2022) and Dang et al. (2023). ...
Many advanced, image-based precision agricultural technologies for plant breeding, field crop research, and site-specific crop management hinge on the reliable detection and phenotyping of plants across highly variable morphological growth stages. Convolutional neural networks (CNNs) have shown promise for image-based plant phenotyping and weed recognition, but their ability to recognize growth stages, often with stark differences in appearance, is uncertain. Amaranthus palmeri (Palmer amaranth) is a particularly challenging weed plant in cotton (Gossypium hirsutum) production, exhibiting highly variable plant morphology both across growth stages over a growing season, as well as between plants at a given growth stage due to high genetic diversity. In this paper, we investigate eight-class growth stage recognition of A. palmeri in cotton as a challenging model for You Only Look Once (YOLO) architectures. We compare 26 different architecture variants from YOLO v3, v5, v6, v6 3.0, v7, and v8 on an eight-class growth stage dataset of A. palmeri. The highest mAP@[0.5:0.95] for recognition of all growth stage classes was 47.34% achieved by v8-X, with inter-class confusion across visually similar growth stages. With all growth stages grouped as a single class, performance increased, with a maximum mean average precision (mAP@[0.5:0.95]) of 67.05% achieved by v7-Original. Single class recall of up to 81.42% was achieved by v5-X, and precision of up to 89.72% was achieved by v8-X. Class activation maps (CAM) were used to understand model attention on the complex dataset. Fewer classes, grouped by visual or size features improved performance over the ground-truth eight-class dataset. Successful growth stage detection highlights the substantial opportunity for improving plant phenotyping and weed recognition technologies with open-source object detection architectures.
... and represent average precision 0.5-0.95 (mAP_0.5:0.95) are considered the most common metrics when evaluating object detectors [25]. ...
The advancement of computer vision technology has allowed for the easy detection of weeds and other stressors in turfgrasses and agriculture. This study aimed to evaluate the feasibility of single shot object detectors for weed detection in lawns, which represents a difficult task. In this study, four different YOLO (You Only Look Once) object detectors version, along with all their various scales, were trained on a public ‘Weeds’ dataset with 4203 digital images of weeds growing in lawns with a total of 11,385 annotations and tested for weed detection in turfgrasses. Different weed species were considered as one class (‘Weeds’). Trained models were tested on the test subset of the ‘Weeds’ dataset and three additional test datasets. Precision (P), recall (R), and mean average precision (mAP_0.5 and mAP_0.5:0.95) were used to evaluate the different model scales. YOLOv8l obtained the overall highest performance in the ‘Weeds’ test subset resulting in a P (0.9476), mAP_0.5 (0.9795), and mAP_0.5:0.95 (0.8123), while best R was obtained from YOLOv5m (0.9663). Despite YOLOv8l high performances, the outcomes obtained on the additional test datasets have underscored the necessity for further enhancements to address the challenges impeding accurate weed detection.
... To address this issue, machine vision-based weed control is emerging as a promising solution, allowing for accurate identification and localization of weed plants and site-specific, individualized treatments such as spot spraying or high-flame laser weed killing. However, the development of robust machine vision systems is heavily reliant on large volumes of labeled image datasets (Westwood et al., 2018;Chen et al., 2022a;Dang et al., 2023), which is often cost-expensive and timeconsuming. As such, there is a growing research interest in developing label-efficient learning algorithms for weed (crop) recognition. ...
The past decade has witnessed many great successes of machine learning (ML) and deep learning (DL) applications in agricultural systems, including weed control, plant disease diagnosis, agricultural robotics, and precision livestock management. Despite tremendous progresses, one downside of such ML/DL models is that they generally rely on large-scale labeled datasets for training, and the performance of such models is strongly influenced by the size and quality of available labeled data samples. In addition, collecting, processing, and labeling such large-scale datasets is extremely costly and time-consuming, partially due to the rising cost in human labor. Therefore, developing label-efficient ML/DL methods for agricultural applications has received significant interests among researchers and practitioners. In fact, there are more than 50 papers on developing and applying deep-learning-based label-efficient techniques to address various agricultural problems since 2016, which motivates the authors to provide a timely and comprehensive review of recent label-efficient ML/DL methods in agricultural applications. To this end, we first develop a principled taxonomy to organize these methods according to the degree of supervision, including weak supervision (i.e., active learning and semi-/weakly- supervised learning), and no supervision (i.e., un-/self- supervised learning), supplemented by representative state-of-the-art label-efficient ML/DL methods. In addition, a systematic review of various agricultural applications exploiting these label-efficient algorithms, such as precision agriculture, plant phenotyping, and postharvest quality assessment, is presented. Finally, we discuss the current problems and challenges, as well as future research directions. A well-classified paper list can be accessed at https://github.com/DongChen06/Label-efficient-in-Agriculture.
... This category of methods is generally computationally efficient but has limitations in terms of localisation when weeds are in close proximity or are occluded by crops. This category of methods make use of Faster RCNN (Jiang et al., 2020) and YOLO family neural networks like YOLO-v3 (Sharpe et al., 2020), YOLOv4 (Zhao et al., 2022), YOLOv5 (Wang et al., 2022), YOLOv6 (Dang et al., 2023), and YOLOv7 (Gallo et al., 2023). These deep learning neural networks are efficient, however, there are two major problems found in their implementation. ...
Agricultural image and vision computing are significantly different from other object classification-based methods because two base classes in agriculture, crops and weeds, have many common traits. Efficient crop, weeds, and soil classification are required to perform autonomous (spraying, harvesting, etc.) activities in agricultural fields. In a three-class (crop–weed–background) agricultural classification scenario, it is usually easier to accurately classify the background class than the crop and weed classes because the background class appears significantly different feature-wise than the crop and weed classes. However, robustly distinguishing between the crop and weed classes is challenging because their appearance features generally look very similar. To address this problem, we propose a framework based on a convolutional W-shaped network with two encoder–decoder structures of different sizes. The first encoder–decoder structure differentiates between background and vegetation (crop and weed), and the second encoder–decoder structure learns discriminating features to classify crop and weed classes efficiently. The proposed W network is generalizable for different crop types. The effectiveness of the proposed network is demonstrated on two crop datasets—a tobacco dataset and a sesame dataset, both collected in this study and made available publicly online for use by the community—by evaluating and comparing the performance with existing related methods. The proposed method consistently outperforms existing related methods on both datasets.
... In addition to the detection of pedestrian crosswalks, object detection processes are also applied in many fields nowadays. Object detection plays an active role in sectors such as health [21], safety [22], transportation [23,24], and agriculture [25,26]. Many researchers enlarge their dataset to achieve high detection accuracy and experiment with differences in the network structure of their detection model. ...
... Detection of weeds and product counting in the agricultural sector provide great opportunities in the marketing part. In a study conducted by [25], the YOLOv7 model reached a mAP value of 61 percent during the detection of weeds in a field. This value might vary depending on the dataset and the detected object. ...
Autonomous vehicles have gained popularity in recent years, but they are still not compatible with other vulnerable components of the traffic system, including pedestrians, bicyclists, motorcyclists, and occupants of smaller vehicles such as passenger cars. This incompatibility leads to reduced system performance and undermines traffic safety and comfort. To address this issue, the authors considered pedestrian crosswalks where vehicles, pedestrians, and micro-mobility vehicles collide at right angles in an urban road network. These road sections are areas where vulnerable people encounter vehicles perpendicularly. In order to prevent accidents in these areas, it is planned to introduce a warning system for vehicles and pedestrians. This procedure consists of multi-stage activities by sending warnings to drivers, disabled individuals, and pedestrians with phone addiction simultaneously. This collective autonomy is expected to reduce the number of accidents drastically. The aim of this paper is the automatic detection of a pedestrian crosswalk in an urban road network, designed from both pedestrian and vehicle perspectives. Faster R-CNN (R101-FPN and X101-FPN) and YOLOv7 network models were used in the analytical process of a dataset collected by the authors. Based on the detection performance comparison between both models, YOLOv7 accuracy was 98.6%, while the accuracy for Faster R-CNN was 98.29%. For the detection of different types of pedestrian crossings, YOLOv7 gave better prediction results than Faster R-CNN, although quite similar results were obtained.
... The popularity is likely the result of ease of use and the high performance achievable on niche datasets (e.g., Dang et al. [2023]). These architectures are typically supplied with pretrained models on large open-source labelled image datasets, such as the COCO (Common Objects in Context) dataset containing over 330,000 images and over 1.5 million object instances. ...
... Without sufficient reporting and oversight of the approach (e.g., data, code, and configuration), reproducibility is diminished. A standardised platform for this critical step may help, where research often appears to be conducted with bespoke implementations of each algorithm (e.g., Dang et al. [2023] and Olsen et al. [2019]). ...
Automating the analysis of plants using image processing would help remove barriers to phenotyping and large-scale precision agricultural technologies, such as site-specific weed control. The combination of accessible hardware and high-performance deep learning (DL) tools for plant analysis is becoming widely recognised as a path forward for both plant science and applied precision agricultural purposes. Yet, a lack of collaboration in image analysis for plant science, despite the open-source origins of much of the technology, is hindering development. Here, we show how tools developed for specific attributes of phenotyping or weed recognition for precision weed control have substantial overlapping data structure, software/hardware requirements and outputs. An open-source approach to these tools facilitates interdisciplinary collaboration, avoiding unnecessary repetition and allowing research groups in both basic and applied sciences to capitalise on advancements and resolve respective bottlenecks. The approach mimics that of machine learning in its nascence. Three areas of collaboration are identified as critical for improving efficiency, (1) standardised, open-source, annotated dataset development with consistent metadata reporting; (2) establishment of accessible and reliable training and testing platforms for DL algorithms; and (3) sharing of all source code used in the research process. The complexity of imaging plants and cost of annotating image datasets means that collaboration from typically distinct fields will be necessary to capitalise on the benefits of DL for both applied and basic science purposes.
Purpose
This paper aims to design an AI-based drone that can facilitate the complicated and time-intensive control process for detecting healthy and defective solar panels. Today, the use of solar panels is becoming widespread, and control problems are increasing. Physical control of the solar panels is critical in obtaining electrical power. Controlling solar panel power plants and rooftop panel applications installed in large areas can be difficult and time-consuming. Therefore, this paper designs a system that aims to panel detection.
Design/methodology/approach
This paper designed a low-cost AI-based unmanned aerial vehicle to reduce the difficulty of the control process. Convolutional neural network based AI models were developed to classify solar panels as damaged, dusty and normal. Two approaches to the solar panel detection model were adopted: Approach 1 and Approach 2.
Findings
The training was conducted with YOLOv5, YOLOv6 and YOLOv8 models in Approach 1. The best F1 score was 81% at 150 epochs with YOLOv5m. In total, 87% and 89% of the best F1 score and mAP values were obtained with the YOLOv5s model at 100 epochs in Approach 2 as a proposed method. The best models at Approaches 1 and 2 were used with a developed AI-based drone in the real-time test application.
Originality/value
The AI-based low-cost solar panel detection drone was developed with an original data set of 1,100 images. A detailed comparative analysis of YOLOv5, YOLOv6 and YOLOv8 models regarding performance metrics was realized. Gaussian, salt-pepper noise addition and wavelet transform noise removal preprocessing techniques were applied to the created data set under the proposed method. The proposed method demonstrated expressive and remarkable performance in panel detection applications.
With the rapid development of artificial intelligence and deep learning technologies, their applications in the field of agriculture, particularly in plant disease detection, have become increasingly extensive. This study focuses on the high-precision detection of tomato diseases, which is of paramount importance for agricultural economic benefits and food safety. To achieve this aim, a tomato disease image dataset was first constructed, and a NanoSegmenter model based on the Transformer structure was proposed. Additionally, lightweight technologies, such as the inverted bottleneck technique, quantization, and sparse attention mechanism, were introduced to optimize the model’s performance and computational efficiency. The experimental results demonstrated excellent performance of the model in tomato disease detection tasks, achieving a precision of 0.98, a recall of 0.97, and an mIoU of 0.95, while the computational efficiency reached an inference speed of 37 FPS. In summary, this study provides an effective solution for high-precision detection of tomato diseases and offers insights and references for future research.
