Journal of Internet of Things and Convergence (사물인터넷융복합논문지)

The Korea Internet of Things Society (한국사물인터넷학회)
권호 표지
DOI QR코드

DOI QR Code

Deep Learning-Based Plant Health State Classification Using Image Data

영상 데이터를 이용한 딥러닝 기반 작물 건강 상태 분류 연구

  • Ali Asgher Syed (Department of Electronics Engineering, Jeonbuk National University) ;
  • Jaehawn Lee (Department of Electronics Engineering, Jeonbuk National University) ;
  • Alvaro Fuentes (Core Research Institute of Intelligent Robots, Jeonbuk National University) ;
  • Sook Yoon (Department of Computer Engineering, Mokpo National University) ;
  • Dong Sun Park (Core Research Institute of Intelligent Robots, Jeonbuk National University)
  • 세이드 알리 에스거 (전북대학교 전자공학부) ;
  • 이재환 (전북대학교 전자공학부) ;
  • 알바로 푸엔테스 (전북대학교 지능형로봇연구소) ;
  • 윤숙 (목포대학교 컴퓨터공학과) ;
  • 박동선 (전북대학교 지능형로봇연구소)
  • Received : 2024.06.28
  • Accepted : 2024.08.16
  • Published : 2024.08.31
Download PDF
⟨ Previous Next ⟩

Abstract

Tomatoes are rich in nutrients like lycopene, β-carotene, and vitamin C. However, they often suffer from biological and environmental stressors, resulting in significant yield losses. Traditional manual plant health assessments are error-prone and inefficient for large-scale production. To address this need, we collected a comprehensive dataset covering the entire life span of tomato plants, annotated across 5 health states from 1 to 5. Our study introduces an Attention-Enhanced DS-ResNet architecture with Channel-wise attention and Grouped convolution, refined with new training techniques. Our model achieved an overall accuracy of 80.2% using 5-fold cross-validation, showcasing its robustness in precisely classifying the health states of tomato plants.

토마토에는 리코펜, β-카로틴 및 비타민 C와 같은 영양소가 풍부하고 세계적으로 많이 소비되는 채소 중 하나이다. 그러나 종종 생물학적 및 환경적 스트레스 요인으로 인해 수확량 손실이 발생한다. 전통적인 작물 건강 평가는 오류가 발생하기 쉽고 대규모 생산에 비효율적이다. 이러한 문제를 해결하기 위해 건강 상태에 대해 1~5로 주석을 메긴 토마토 전체 생육기간을 다루는 포괄적인 데이터 세트를 수집하였다. 우리는 Channel-wise attention과 Grouped convolution을 사용한 Attention-Enhanced DS-ResNet 아키텍처와 새로운 학습 기법을 제안한다. 우리의 모델은 5-fold 교차 검증을 사용하여 전체 정확도 80.2%를 달성하여 작물의 건강 상태를 정확하게 분류하는데 있어 견고성을 보여주었다.

Keywords

Acknowledgement

본 결과물은 농림축산식품부 및 과학기술정보통신부, 농촌진흥청의 재원으로 농림식품기술기획평가원과 재단법인 스마트팜연구개발사업단의 스마트팜다부처패키지혁신기술개발사업의 지원을 받아 연구되었음 (RS-2021-IP421005).

References

  1. A. Gerszberg, K. Hnatuszko-Konka, T. Kowalczyk, and A. K. Kononowicz, "Tomato (Solanum lycopersicum L.) in the service of biotechnology," Plant Cell, Tissue and Organ Culture (PCTOC), Vol.120, No.3, pp.881-902, 2015.
  2. V. Bergougnoux, "The history of tomato: From domestication to biopharming," Biotechnol Adv, Vol.32, No.1, pp.170-189, 2014.
  3. A. Fuentes, Y. Lee, Y. Hong, and S. Yoon, "Characteristics of Tomato Plant Diseases - A Study for Tomato Plant Disease Identification," in ISITC 2016 International Symposium on Information Technology Convergence, Shanghai, China, 2016.
  4. Food and Agriculture Organization of the United Nations, "fao_website (https://www.fao.org/faostat/en/#data/QCL)," 2024.
  5. M. Nawaz et al., "A review of plants strategies to resist biotic and abiotic environmental stressors," Science of The Total Environment, Vol.900, p.165832, 2023.
  6. G. K. Sandhu and R. Kaur, "Plant Disease Detection Techniques: A Review," in 2019 International Conference on Automation, Computational and Technology Management (ICACTM), IEEE, pp.34-38, 2019.
  7. M. Brahimi, K. Boukhalfa, and A. Moussaoui, "Deep Learning for Tomato Diseases: Classification and Symptoms Visualization," Applied Artificial Intelligence, Vol.31, No.4, pp.299-315, 2017.
  8. Y. Tian, E. Li, Z. Liang, M. Tan, and X. He, "Diagnosis of Typical Apple Diseases: A Deep Learning Method Based on Multi-Scale Dense Classification Network," Front Plant Sci, Vol.12, Oct. 2021.
  9. J. L. de Moraes, J. de Oliveira Neto, C. Badue, T. Oliveira-Santos, and A. F. de Souza, "Yolo-Papaya: A Papaya Fruit Disease Detector and Classifier Using CNNs and Convolutional Block Attention Modules," Electronics (Basel), Vol.12, No.10, p.2202, 2023.
  10. A. Fuentes, S. Yoon, J. Lee, and D. S. Park, "High-Performance Deep Neural Network-Based Tomato Plant Diseases and Pests Diagnosis System With Refinement Filter Bank," Front Plant Sci, Vol.9, 2018.
  11. A. Guerrero-Ibanez and A. Reyes-Munoz, "Monitoring Tomato Leaf Disease through Convolutional Neural Networks," Electronics (Basel), Vol.12, No.1, p.229, 2023.
  12. J. Dong, A. Fuentes, S. Yoon, H. Kim, and D. S. Park, "An iterative noisy annotation correction model for robust plant disease detection," Front Plant Sci, Vol.14, 2023.
  13. Y. Meng, M. Xu, H. Kim, S. Yoon, Y. Jeong, and D. S. Park, "Known and unknown class recognition on plant species and diseases," Comput Electron Agric, Vol.215, p.108408, Dec. 2023.
  14. F. Arshad et al., "PLDPNet: End-to-end hybrid deep learning framework for potato leaf disease prediction," Alexandria Engineering Journal, Vol.78, pp.406-418, 2023.
  15. W. Xie, S. Wei, Z. Zheng, Y. Jiang, and D. Yang, "Recognition of Defective Carrots Based on Deep Learning and Transfer Learning," Food Bioproc Tech, Vol.14, No.7, pp.1361-1374, 2021.
  16. K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, pp.770-778. 2016.
  17. J. Hu, L. Shen, and G. Sun, "Squeeze-and-Excitation Networks," in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE, pp.7132-7141, 2018.
  18. A. G. Howard et al., "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications," CoRR, Vol.abs/1704.04861, 2017.