Design and Development of a Smart Variable Rate Sprayer Using Deep Learning

Hussain, Nazar; Farooque, Aitazaz A.; Schumann, Arnold W.; McKenzie‐Gopsill, Andrew; Esau, Travis; Abbas, Farhat; Acharya, Bishnu; Zaman, Qamar

doi:10.3390/rs12244091

Cited by 45 publications

(25 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, using disease information for guiding VA has been a difficult problem for precise application owing to the difficulty of disease spot information acquisition. [39][40][41] Ma et al proposed a DCNN for symptom recognition of four diseases in cucumbers, namely anthracnose, downy mildew, powdery mildew, and target leaf spot, with an accuracy of 93.4%. 42 Song et al proposed a new Google net initial module-based apple disease recognition model that is capable of rapid detection and classification of diseases in apples, including rust, spotted leaves, and anthracnose, with a recognition accuracy of 98.5%.…”

Section: Discussionmentioning

confidence: 99%

Variable rate air‐assisted spray based on real‐time disease spot identification

Zhang

Yin

Zhou

et al. 2022

Pest Management Science

View full text Add to dashboard Cite

Background Currently, the variable‐rate application (VA) of agrochemicals on fruit trees is based on canopy volume and biomass. The canopy volume has a significant relationship with disease resistance and degree of disease incidence. Therefore, this study proposes a VA method that uses deep convolutional neural networks for real‐time recognition of disease spots on pear trees. Furthermore, it specifies the limitations and application scenarios of the disease spot recognition. Field performance tests were conducted to verify the performance of the proposed VA system. Results The results showed a mean average precision, precision, and recall of 87.42%, 83.76%, and 87.23%, respectively. The spot recognition rate was 81.3% when the canopy sampling distance, spot diameter, and canopy porosity were 1.2 m, 4–8 mm, and 55.76%, respectively. The results indicate that the proposed VA system saved 51.9% spray volume compared to conventional methods while ensuring quality. Conclusion Compared to the traditional constant rate model, the proposed VA technology based on real‐time disease spot identification can reduce spraying in nondiseased areas, thereby abandoning the previous saturation application practice and significantly reducing pesticide use. © 2022 Society of Chemical Industry.

show abstract

Section: Discussionmentioning

confidence: 99%

Variable rate air‐assisted spray based on real‐time disease spot identification

Zhang

Yin

Zhou

et al. 2022

Pest Management Science

View full text Add to dashboard Cite

show abstract

“…A Canon EOS Rebel T6 DSLR camera ("Canon T6"), an LG G6-H873 smartphone ("LG G6"), and a Logitech c920 HD Pro USB 2.0 webcam ("Logitech c920") were used to capture images at each test plot. The Logitech c920 was selected for its low cost and successful deployment in smart variable-rate sprayers developed by [12] and [13]. The Logitech c920 was mounted to a tripod and connected to a USB 3.1 port on an MSI workstation laptop (WS65 9TM-1410CA, Micro-Star International Co., Ltd) with an Nvidia Quadro RTX 5000 graphics processing unit (GPU) via a 2 m USB 3.0 extension cable.…”

Section: A Field Image Collectionmentioning

confidence: 99%

“…These smart sprayers have typically attached cameras 1.1 m [7], [8] to 1.2 m [9]- [11] from the ground on the applicator boom ahead of the spray nozzles. This boom height is higher than smart sprayers in other cropping systems [12]- [15], but is necessary due to the highly variable topography of wild blueberry fields (Figure 3).…”

Section: Introductionmentioning

confidence: 97%

Parameters Influencing Hair Fescue And Sheep Sorrel Identification In Wild Blueberry Fields Using Convolutional Neural Networks

Hennessy¹,

Esau²,

Zaman³

et al. 2021

Progress in Canadian Mechanical Engineering. Volume 4

Self Cite

View full text Add to dashboard Cite

Broadcast applications of liquid herbicides are used to manage weeds such as hair fescue (Festuca filiformis Pourr.) and sheep sorrel (Rumex acetosella L.) in wild blueberry (Vaccinium angustifolium Ait.) fields. Weeds typically grow in patches of the fields, consequently, herbicide is wasted on areas of the field without weed cover. Application efficiency can be optimized by employing a smart sprayer which uses machine vision to identify areas of the field containing the target weeds in real-time. The YOLOv3-Tiny convolutional neural network (CNN) was trained to detect hair fescue and sheep sorrel using 1280x720 resolution images captured in 58 wild blueberry fields throughout Nova Scotia, Canada. The trained CNN detected at least one target weed per validation image with F1-scores of 0.97 for hair fescue and 0.90 for sheep sorrel at a network resolution of 1280x736. An evaluation was performed at a commercial wild blueberry field in Debert, Nova Scotia, to examine the effects of camera selection and target distance on detection accuracy. A Logitech c920 webcam, an LG G6 smartphone, and a Canon T6 DSLR camera were used to capture colour images at distances of 0.57 m, 0.98 m, and 1.29 m from target weeds. Test plots were selected at randomly spaced intervals along an inverted "W" pattern in the field. Mean F1-scores for each combination of camera and image height were analyzed in a 3x3 factorial arrangement for hair fescue and a 3x2 factorial arrangement for sheep sorrel. The peak F1-score for detection of at least one hair fescue plant, 0.97, was achieved with images captured with the LG G6 smartphone at a height of 0.98 m. Images captured with the LG G6 smartphone and Canon T6 DSLR camera at 0.98 m each achieved an F1-score of 0.82 for detection of at least one sheep sorrel plant per image. Sheep sorrel was only detected by the CNN in images from the Logitech c920 camera using 3 of 9 parameter combinations in the analysis. Future work will examine images from two additional fields tested under similar conditions. Additionally, the CNN will be used to control herbicide applications after integration with a real-time smart sprayer. A web-based application will be developed to detect target weeds using the CNN and provide wild blueberry growers with site-specific information to aid management decisions. Using a CNN to detect weeds will improve traditional management techniques and create cost-savings and greater sustainability for wild blueberry growers.

show abstract

“…The most common use of CNNs in the agricultural sector involves image analysis; CNNs analyze the textural, spectral, and spatial features of images and can extract features unseen by the human eye (Albawi et al 2017;Sapkota et al 2020). Fruit counting, weed detection, disease detection, and grain yield estimation are ways that CNNs have been used in agriculture (Biffi et al 2021;Hussain et al 2020Hussain et al , 2021Sivakumar et al 2020;Yang et al 2019).…”

Section: Introductionmentioning

confidence: 99%

Use of open-source object detection algorithms to detect Palmer amaranth (Amaranthus palmeri) in soybean

et al. 2022

View full text Add to dashboard Cite

Site-specific weed management using open-sourced object detection algorithms could accurately detect weeds in cropping systems. We investigated the use of object detection algorithms to detect Palmer amaranth (Amaranthus palmeri S. Watson) in soybean [Glycine max (L.) Merr.]. The objectives were to 1) develop an annotated image database of A. palmeri and soybean to fine-tune object detection algorithms, 2) compare effectiveness of multiple open-sourced algorithms in detecting A. palmeri, and 3) evaluate the relationship between A. palmeri growth features and A. palmeri detection ability. Soybean field sites were established in Manhattan, KS and Gypsum, KS with natural populations of A. palmeri. A total of 1108 and 392 images were taken aerially and at ground level, respectively, between May 27 and July 27, 2021. After image annotation, a total of 4492 images were selected. Annotated images were used to fine tune open-source Faster regional convolutional (Faster R-CNN) and Single Shot Detector (SSD) algorithms using a Resnet backbone, as well as the You Only Look Once (YOLO) series algorithms. Results demonstrated that YOLO version 5 achieved the highest mean average precision score of 0.77. For both A. palmeri and soybean detections within this algorithm, the highest F1 score was 0.72 when using a confidence threshold of 0.298. A lower confidence threshold of 0.15 increased the likelihood of species detection, but also increased the likelihood of false positive detections. The trained YOLOv5 dataset was used to identify A. palmeri in a dataset paired with measured growth features. Linear regression models predicted that as A. palmeri densities increased and as A. palmeri height increased, precision, recall, and F1 scores of algorithms decreased. We conclude that open-sourced algorithms such as YOLOv5 show great potential in detecting A. palmeri in soybean cropping systems.

show abstract

Design and Development of a Smart Variable Rate Sprayer Using Deep Learning

Cited by 45 publications

References 30 publications

Variable rate air‐assisted spray based on real‐time disease spot identification

Variable rate air‐assisted spray based on real‐time disease spot identification

Parameters Influencing Hair Fescue And Sheep Sorrel Identification In Wild Blueberry Fields Using Convolutional Neural Networks

Use of open-source object detection algorithms to detect Palmer amaranth (Amaranthus palmeri) in soybean

Contact Info

Product

Resources

About