A Deep Learning Based Approach for Identifying Whitefly Crop Pests for Precision Agriculture Applications

This paper presents a machine learning and deep learning model for detection of white fly infestation. The proposed scheme presents two network architectures v.i.z. the Deep Bayes Net, the convolutional neural network, the ResNet and the MIL-CNN for the classification of images. The proposed model presents a rigorous image pre-processing paradigm prior to feature computation, which makes the system more robust and immune to noise effects. The statistical feature extraction presented in this paper would be applicable to a wide range of images thereby rendering flexibility to the choice of data source and image formats. The performance metrics of the proposed system have been chosen as accuracy, sensitivity, specificity and precision. The proposed MIL-CNN outperforms the other models in terms of accuracy. As manual labelling the data for large datasets is extremely cumbersome and prone to errors, hence further directions of research can be exploring self supervised learning (SSL) models and techniques to reduce the complexity in labelling large datasets. Future directions of research in the domain can explore the utility and efficacy of transfer learning. This would allow pre-trained models to be tested on new diverse datasets which may reduce the time spent in labelling of new datasets and make use of the existing pre-trained models for time critical applications of machine learning in the medical field.

[1] W Li, D Wang, M Li, Y Gao, J Wu, X Yang, “Field detection of tiny pests from sticky trap images using deep learning in agricultural greenhouse”, Computers and Electronics in Agriculture, Elsevier, 2025 Volume 183, pp.106048-106059. [2] KRB Legaspi, NWS Sison, “Detection and Classification of Whiteflies and Fruit Flies Using YOLO”, 13th International Conference on Computer and Automation Engineering (ICCAE), IEEE 2024, pp. 1-4 [3] G Pattnaik, K Parvathi, “Automatic Detection and Classification of Tomato Pests Using Support Vector Machine Based on HOG and LBP Feature Extraction Technique”, Progress in Advanced Computing and Intelligent Engineering. Advances in Intelligent Systems and Computing, Springer 2023, Volume- 1199, pp.49-55. [4] G Pattnaik, VK Shrivastava, “Transfer Learning-Based Framework for Classification of Pest in Tomato Plants”, Journal of Applied Artificial Intelligence, Taylor and Francis, 2022, Volume-34, Issue-13, pp. 981-993. [5] QJ Wang, SY Zhang, SF Dong, GC Zhang, “Pest24: A large-scale very small object data set of agricultural pests for multi-target detection”, Computers and Electronics in Agriculture, Elsevier 2021, Volume-175, pp. 105585 [6] DJA Rustia, CE Lin, JY Chung, YJ Zhuang, “Application of an image and environmental sensor network for automated greenhouse insect pest monitoring”, Journal of Asia Pacific Entomology, Elsevier 2020, Volume-23, Issue-1, pp.17-28. [7] L Liu, R Wang, C Xie, P Yang, F Wang, “PestNet: An end-to-end deep learning approach for large-scale multi-class pest detection and classification”, IEEE Access 2019, Volume-7. pp. 45301-45312. [8] CJ Chen, JS Wu, CY Chang, YM Huang, “Agricultural pests damage detection using deep learning”, Advances in Intelligent Systems and Computing, Springer 2018, Volume-1036, pp.545-554. [9] L Deng, Y Wang, Z Han, R Yu, “Research on insect pest image detection and recognition based on bio-inspired methods”, Journal of Biosystems Engineering, Elsevier 2018, Volume 169, pp. 139-148 [10] MA Ebrahimi, MH Khoshtaghaza, S Minaei, “Vision-based pest detection based on SVM classification method”, Journal of Computers and Electronics in Agriculture, Elsevier 2017, Volume 137, pp. 52-58