Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection
Tomato diseases, including Leaf blight, Leaf curl, Septoria leaf spot, and Verticillium wilt, are responsible for up to 50% of annual yield loss, significantly impacting global tomato production, valued at approximately USD 87 billion. In Ghana, there is a yield gap of about 50% in tomato production...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-09-01
|
| Series: | AgriEngineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2624-7402/6/4/203 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846106428549365760 |
|---|---|
| author | Dennis Agyemanh Nana Gookyi Fortunatus Aabangbio Wulnye Michael Wilson Paul Danquah Samuel Akwasi Danso Awudu Amadu Gariba |
| author_facet | Dennis Agyemanh Nana Gookyi Fortunatus Aabangbio Wulnye Michael Wilson Paul Danquah Samuel Akwasi Danso Awudu Amadu Gariba |
| author_sort | Dennis Agyemanh Nana Gookyi |
| collection | DOAJ |
| description | Tomato diseases, including Leaf blight, Leaf curl, Septoria leaf spot, and Verticillium wilt, are responsible for up to 50% of annual yield loss, significantly impacting global tomato production, valued at approximately USD 87 billion. In Ghana, there is a yield gap of about 50% in tomato production, which requires drastic measures to increase the yield of tomatoes. Conventional diagnostic methods are labor-intensive and impractical for real-time application, highlighting the need for innovative solutions. This study addresses these issues in Ghana by utilizing Edge Impulse to deploy multiple deep-learning models on a single mobile device, facilitating the rapid and precise detection of tomato leaf diseases in the field. This work compiled and rigorously prepared a comprehensive Ghanaian dataset of tomato leaf images, applying advanced preprocessing and augmentation techniques to enhance robustness. Using TensorFlow, we designed and optimized efficient convolutional neural network (CNN) architectures, including MobileNet, Inception, ShuffleNet, Squeezenet, EfficientNet, and a custom Deep Neural Network (DNN). The models were converted to TensorFlow Lite format and quantized to int8, substantially reducing the model size and improving inference speed. Deployment files were generated, and the Edge Impulse platform was configured to enable multiple model deployments on a mobile device. Performance evaluations across edge hardware provided metrics such as inference speed, accuracy, and resource utilization, demonstrating reliable real-time detection. EfficientNet achieved a high training accuracy of 97.12% with a compact 4.60 MB model size, proving its efficacy for mobile device deployment. In contrast, the custom DNN model is optimized for microcontroller unit (MCU) deployment. This edge artificial intelligence (AI) technology integration into agricultural practices offers scalable, cost-effective, and accessible solutions for disease classification, enhancing crop management, and supporting sustainable farming practices. |
| format | Article |
| id | doaj-art-8f3bec2429584573a7eeac3b461f2527 |
| institution | Kabale University |
| issn | 2624-7402 |
| language | English |
| publishDate | 2024-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | AgriEngineering |
| spelling | doaj-art-8f3bec2429584573a7eeac3b461f25272024-12-27T14:03:32ZengMDPI AGAgriEngineering2624-74022024-09-01643563358510.3390/agriengineering6040203Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease DetectionDennis Agyemanh Nana Gookyi0Fortunatus Aabangbio Wulnye1Michael Wilson2Paul Danquah3Samuel Akwasi Danso4Awudu Amadu Gariba5Electronics Division, Institute for Scientific and Technological Information, Council for Scientific and Industrial Research, Accra CT-2211, GhanaDepartment of Telecommunication Engineering, Kwame Nkrumah University of Science and Technology, Kumasi AK-509-4752, GhanaElectronics Division, Institute for Scientific and Technological Information, Council for Scientific and Industrial Research, Accra CT-2211, GhanaElectronics Division, Institute for Scientific and Technological Information, Council for Scientific and Industrial Research, Accra CT-2211, GhanaDepartment of Computer Engineering, Ghana Communication Technology University, Accra PMB-100, GhanaPlant Protection and Regulatory Services Directorate, The Ministry of Food and Agriculture, Accra M-37, GhanaTomato diseases, including Leaf blight, Leaf curl, Septoria leaf spot, and Verticillium wilt, are responsible for up to 50% of annual yield loss, significantly impacting global tomato production, valued at approximately USD 87 billion. In Ghana, there is a yield gap of about 50% in tomato production, which requires drastic measures to increase the yield of tomatoes. Conventional diagnostic methods are labor-intensive and impractical for real-time application, highlighting the need for innovative solutions. This study addresses these issues in Ghana by utilizing Edge Impulse to deploy multiple deep-learning models on a single mobile device, facilitating the rapid and precise detection of tomato leaf diseases in the field. This work compiled and rigorously prepared a comprehensive Ghanaian dataset of tomato leaf images, applying advanced preprocessing and augmentation techniques to enhance robustness. Using TensorFlow, we designed and optimized efficient convolutional neural network (CNN) architectures, including MobileNet, Inception, ShuffleNet, Squeezenet, EfficientNet, and a custom Deep Neural Network (DNN). The models were converted to TensorFlow Lite format and quantized to int8, substantially reducing the model size and improving inference speed. Deployment files were generated, and the Edge Impulse platform was configured to enable multiple model deployments on a mobile device. Performance evaluations across edge hardware provided metrics such as inference speed, accuracy, and resource utilization, demonstrating reliable real-time detection. EfficientNet achieved a high training accuracy of 97.12% with a compact 4.60 MB model size, proving its efficacy for mobile device deployment. In contrast, the custom DNN model is optimized for microcontroller unit (MCU) deployment. This edge artificial intelligence (AI) technology integration into agricultural practices offers scalable, cost-effective, and accessible solutions for disease classification, enhancing crop management, and supporting sustainable farming practices.https://www.mdpi.com/2624-7402/6/4/203Edge ImpulseTensorFlowCNNtomato disease detectionagriculturemicrocontroller unit |
| spellingShingle | Dennis Agyemanh Nana Gookyi Fortunatus Aabangbio Wulnye Michael Wilson Paul Danquah Samuel Akwasi Danso Awudu Amadu Gariba Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection AgriEngineering Edge Impulse TensorFlow CNN tomato disease detection agriculture microcontroller unit |
| title | Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection |
| title_full | Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection |
| title_fullStr | Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection |
| title_full_unstemmed | Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection |
| title_short | Enabling Intelligence on the Edge: Leveraging Edge Impulse to Deploy Multiple Deep Learning Models on Edge Devices for Tomato Leaf Disease Detection |
| title_sort | enabling intelligence on the edge leveraging edge impulse to deploy multiple deep learning models on edge devices for tomato leaf disease detection |
| topic | Edge Impulse TensorFlow CNN tomato disease detection agriculture microcontroller unit |
| url | https://www.mdpi.com/2624-7402/6/4/203 |
| work_keys_str_mv | AT dennisagyemanhnanagookyi enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection AT fortunatusaabangbiowulnye enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection AT michaelwilson enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection AT pauldanquah enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection AT samuelakwasidanso enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection AT awuduamadugariba enablingintelligenceontheedgeleveragingedgeimpulsetodeploymultipledeeplearningmodelsonedgedevicesfortomatoleafdiseasedetection |