Edge-AI in Precision Agriculture: A Quantized MobileNetV2 Framework for Localized Crop Disease Detection in Punjab, Pakistan
DOI:
https://doi.org/10.36755/jac.v4i1.142Keywords:
edge artificial intelligence; , precision agriculture; , MobileNetV2; , integer quantization; , crop disease classification; , domain shift; , wheat rust; , cotton leaf curl disease.Abstract
In rural Punjab, crop disease recognition needs models that can deal with the changeability of field images and work without a consistent network connection. High accuracy on controlled leaf-image datasets does not form performance on locally grown wheat and cotton, while cloud dependent inference introduces communication and service-availability constraints. In this work, an offline-first framework with localized image classification, MobileNetV2 backbone and integer quantization for smartphone deployment is proposed. The initial dataset inventory includes 10,000 images from the agricultural areas of Faisalabad (2,000 healthy wheat, 2,000 rust infected wheat, 2,000 healthy cotton, 2,500 cotton leaf curl and 1,500 cotton bacterial blight images). The proposed pipeline is built around 224 × 224 RGB inputs, crop-aware interpretation and a five-class softmax output, and training-only augmentation. A validation protocol is outlined that can be replicated on farms for each farm group separation, field testing and device-level benchmarking. The reported aggregate accuracy is 94.8% for the floating point MobileNetV2 model and 94.5% after quantization; the model size is reduced from 14 to ~3.5 MB. These values represent 75% reduction in storage and 0.3 percentage point accuracy loss. The highest reported accuracy of 96.3% is achieved by ResNet50 with a significantly higher reported storage. The quantized latency is only reported as less than 200ms, so it is not known whether it is faster than the 180ms floating-point model. What it is is a locally scoped deployment and evaluation framework, not a new convolutional architecture. Evidence is supportive of preliminary storage-efficient symptom screening, but not proven presymptomatic detection, cross-district generalization or reduced pesticide use. To support an operational rollout, class-specific validation of prediction-level records needs to be demonstrated as well as verified model artifacts and future farmer testing.
Downloads
References
1. Savary S, Willocquet L, Pethybridge SJ, Esker P, McRoberts N, Nelson A. The global burden of pathogens and pests on major food crops. Nat Ecol Evol. 2019;3:430–439. doi: 10.1038/s41559-018-0793-y. DOI: https://doi.org/10.1038/s41559-018-0793-y
2. Benos L, Tagarakis AC, Dolias G, Berruto R, Kateris D, Bochtis D. Machine Learning in Agriculture: A Comprehensive Updated Review. Sensors. 2021;21:3758. doi: 10.3390/s21113758. DOI: https://doi.org/10.3390/s21113758
3. Shoaib M, et al. An advanced deep learning models-based plant disease detection: A review of recent research. Front Plant Sci. 2023;14:1158933. doi: 10.3389/fpls.2023.1158933. DOI: https://doi.org/10.3389/fpls.2023.1158933
4. Upadhyay A, Chandel NS, Singh KP, Chakraborty SK, Nandede BM, Kumar M, et al. Deep learning and computer vision in plant disease detection: a comprehensive review of techniques, models, and trends in precision agriculture. Artif Intell Rev. 2025;58:92. doi: 10.1007/s10462-024-11100-x. DOI: https://doi.org/10.1007/s10462-024-11100-x
5. Nazeer R, Ali S, Hu Z, Ansari GJ, Al-Razgan M, Awwad EM, et al. Detection of cotton leaf curl disease’s susceptibility scale level based on deep learning. J Cloud Comput. 2024;13:50. doi: 10.1186/s13677-023-00582-9. DOI: https://doi.org/10.1186/s13677-023-00582-9
6. Ahmad N, Qadri S, Akhtar N. A multimodal fusion framework to diagnose cotton leaf curl virus using machine vision techniques. Cogent Food Agric. 2024;10:2339572. doi: 10.1080/23311932.2024.2339572. DOI: https://doi.org/10.1080/23311932.2024.2339572
7. Nigam S, et al. Deep transfer learning model for disease identification in wheat crop. Ecol Inform. 2023;75:102068. doi: 10.1016/j.ecoinf.2023.102068. DOI: https://doi.org/10.1016/j.ecoinf.2023.102068
8. Mi Z, Zhang X, Su J, Han D, Su B. Wheat stripe rust grading by deep learning with attention mechanism and images from mobile devices. Front Plant Sci. 2020;11:558126. doi: 10.3389/fpls.2020.558126. DOI: https://doi.org/10.3389/fpls.2020.558126
9. Barbedo JGA. A review on the main challenges in automatic plant disease identification based on visible range images. Biosyst Eng. 2016;144:52–60. doi: 10.1016/j.biosystemseng.2016.01.017. DOI: https://doi.org/10.1016/j.biosystemseng.2016.01.017
10. Mohanty SP, Hughes DP, Salathé M. Using deep learning for image-based plant disease detection. Front Plant Sci. 2016;7:1419. doi: 10.3389/fpls.2016.01419. DOI: https://doi.org/10.3389/fpls.2016.01419
11. Ferentinos KP. Deep learning models for plant disease detection and diagnosis. Comput Electron Agric. 2018;145:311–318. doi: 10.1016/j.compag.2018.01.009. DOI: https://doi.org/10.1016/j.compag.2018.01.009
12. Moupojou E, et al. FieldPlant: A Dataset of Field Plant Images for Plant Disease Detection and Classification With Deep Learning. IEEE Access. 2023;11:35398–35410. doi: 10.1109/ACCESS.2023.3263042. DOI: https://doi.org/10.1109/ACCESS.2023.3263042
13. Wu X, et al. From Laboratory to Field: Unsupervised Domain Adaptation for Plant Disease Recognition in the Wild. Plant Phenomics. 2023;5:0038. doi: 10.34133/plantphenomics.0038. DOI: https://doi.org/10.34133/plantphenomics.0038
14. Yang S, Feng Q, Zhang J, Yang W, Zhou W, Yan W. From laboratory to field: cross-domain few-shot learning for crop disease identification in the field. Front Plant Sci. 2024;15:1434222. doi: 10.3389/fpls.2024.1434222. DOI: https://doi.org/10.3389/fpls.2024.1434222
15. Elfatimi E, Eryiğit R, Shehu HA. Impact of datasets on the effectiveness of MobileNet for beans leaf disease detection. Neural Comput Appl. 2024;36:1773–1789. doi: 10.1007/s00521-023-09187-4. DOI: https://doi.org/10.1007/s00521-023-09187-4
16. Khan AT, Jensen SM, Khan AR, Li S. Plant disease detection model for edge computing devices. Front Plant Sci. 2023;14:1308528. doi: 10.3389/fpls.2023.1308528. DOI: https://doi.org/10.3389/fpls.2023.1308528
17. Zhou H, Chen J, Niu X, Dai Z, Qin L, Ma L, et al. Identification of leaf diseases in field crops based on improved ShuffleNetV2. Front Plant Sci. 2024;15:1342123. doi: 10.3389/fpls.2024.1342123. DOI: https://doi.org/10.3389/fpls.2024.1342123
18. Mazumder MKA, Mridha MF, Alfarhood S, Safran M, Abdullah-Al-Jubair M, Che D. A robust and light-weight transfer learning-based architecture for accurate detection of leaf diseases across multiple plants using less amount of images. Front Plant Sci. 2024;14:1321877. doi: 10.3389/fpls.2023.1321877. DOI: https://doi.org/10.3389/fpls.2023.1321877
19. Quan S, Wang J, Jia Z, Yang M, Xu Q. MS-Net: a novel lightweight and precise model for plant disease identification. Front Plant Sci. 2023;14:1276728. doi: 10.3389/fpls.2023.1276728. DOI: https://doi.org/10.3389/fpls.2023.1276728
20. Xu Y, Li D, Li C, Yuan Z, Dai Z. LiSA-MobileNetV2: an extremely lightweight deep learning model with Swish activation and attention mechanism for accurate rice disease classification. Front Plant Sci. 2025;16:1619365. doi: 10.3389/fpls.2025.1619365. DOI: https://doi.org/10.3389/fpls.2025.1619365
21. Ugwah VN, Abolghasemi V. Optimised MobileNet for very lightweight and accurate plant leaf disease detection. Sci Rep. 2025;15:43690. doi: 10.1038/s41598-025-27393-z. DOI: https://doi.org/10.1038/s41598-025-27393-z
22. Pal C, Karmakar S, Mukherjee I, et al. A lightweight and explainable CNN model for empowering plant disease diagnosis. Sci Rep. 2025;15:30720. doi: 10.1038/s41598-025-94083-1. DOI: https://doi.org/10.1038/s41598-025-94083-1
23. Bi C, Xu S, Hu N, Zhang S, Zhu Z, Yu H. Identification method of corn leaf disease based on improved MobileNetV3 model. Agronomy. 2023;13:300. doi: 10.3390/agronomy13020300. DOI: https://doi.org/10.3390/agronomy13020300
24. Yang L, Yu X, Zhang S, Long H, Zhang H, Xu S, et al. GoogLeNet based on residual network and attention mechanism identification of rice leaf diseases. Comput Electron Agric. 2023;204:107543. doi: 10.1016/j.compag.2022.107543. DOI: https://doi.org/10.1016/j.compag.2022.107543
25. Bhagat M, Kumar D, Kumar S. Bell pepper leaf disease classification with LBP and VGG-16 based fused features and RF classifier. Int J Inf Technol. 2023;15:465–475. doi: 10.1007/s41870-022-01136-z. DOI: https://doi.org/10.1007/s41870-022-01136-z
26. Memon MS, Kumar P, Iqbal R. Meta Deep Learn Leaf Disease Identification Model for Cotton Crop. Computers. 2022;11:102. doi: 10.3390/computers11070102. DOI: https://doi.org/10.3390/computers11070102
27. Tao Y, Chang F, Huang Y, Ma L, Xie L, Su H. Cotton Disease Detection Based on ConvNeXt and Attention Mechanisms. IEEE J Radio Freq Identif. 2022;6:805–809. doi: 10.1109/JRFID.2022.3206841. DOI: https://doi.org/10.1109/JRFID.2022.3206841
28. Bao W, Cheng T, Zhou XG, Guo W, Wang Y, Zhang X, et al. An improved DenseNet model to classify the damage caused by cotton aphid. Comput Electron Agric. 2022;203:107485. doi: 10.1016/j.compag.2022.107485. DOI: https://doi.org/10.1016/j.compag.2022.107485
29. Chang S, et al. Recognition of wheat rusts in a field environment based on improved DenseNet. Biosyst Eng. 2024;238:10–21. doi: 10.1016/j.biosystemseng.2023.12.016. DOI: https://doi.org/10.1016/j.biosystemseng.2023.12.016
30. Liu W, Chen Y, et al. StripeRust-Pocket: A Mobile-Based Deep Learning Application for Efficient Disease Severity Assessment of Wheat Stripe Rust. Plant Phenomics. 2024;6:0201. doi: 10.34133/plantphenomics.0201. DOI: https://doi.org/10.34133/plantphenomics.0201
31. Singh R, Krishnan P, Singh VK, Banerjee K. Application of thermal and visible imaging to estimate stripe rust disease severity in wheat using supervised image classification methods. Ecol Inform. 2022;71:101774. doi: 10.1016/j.ecoinf.2022.101774. DOI: https://doi.org/10.1016/j.ecoinf.2022.101774
32. Li Y, et al. Semantic Segmentation of Wheat Stripe Rust Images Using Deep Learning. Agronomy. 2022;12:2933. doi: 10.3390/agronomy12122933. DOI: https://doi.org/10.3390/agronomy12122933
33. Nanavaty A, et al. Integrating deep learning for visual question answering in Agricultural Disease Diagnostics: Case Study of Wheat Rust. Sci Rep. 2024;14:28203. doi: 10.1038/s41598-024-79793-2. DOI: https://doi.org/10.1038/s41598-024-79793-2
34. Argüeso D, Picon A, Irusta U, Medela A, San-Emeterio MG, Bereciartua A, et al. Few-Shot Learning approach for plant disease classification using images taken in the field. Comput Electron Agric. 2020;175:105542. doi: 10.1016/j.compag.2020.105542. DOI: https://doi.org/10.1016/j.compag.2020.105542
35. Borhani Y, Khoramdel J, Najafi E. A deep learning based approach for automated plant disease classification using vision transformer. Sci Rep. 2022;12:11554. doi: 10.1038/s41598-022-15163-0. DOI: https://doi.org/10.1038/s41598-022-15163-0
36. Sambana B, Nnadi HS, Wajid MA, Fidelia NO, Camacho-Zuñiga C, Ajuzie HD, et al. An efficient plant disease detection using transfer learning approach. Sci Rep. 2025;15:19082. doi: 10.1038/s41598-025-02271-w. DOI: https://doi.org/10.1038/s41598-025-02271-w
37. Jackulin C, Murugavalli S. A comprehensive review on detection of plant disease using machine learning and deep learning approaches. Meas Sens. 2022;24:100441. doi: 10.1016/j.measen.2022.100441. DOI: https://doi.org/10.1016/j.measen.2022.100441
38. Arsenovic M, Karanovic M, Sladojevic S, Anderla A, Stefanovic D. Solving Current Limitations of Deep Learning Based Approaches for Plant Disease Detection. Plants. 2019;8:163. doi: 10.3390/plants8060163. DOI: https://doi.org/10.3390/sym11070939
39. Too EC, Yujian L, Njuki S, Yingchun L. A comparative study of fine-tuning deep learning models for plant disease identification. Comput Electron Agric. 2019;161:272–279. doi: 10.1016/j.compag.2018.03.032. DOI: https://doi.org/10.1016/j.compag.2018.03.032
40. Lee SH, Goëau H, Bonnet P, Joly A. New perspectives on plant disease characterization based on deep learning. Comput Electron Agric. 2020;170:105220. doi: 10.1016/j.compag.2020.105220. DOI: https://doi.org/10.1016/j.compag.2020.105220
41. Kabala DM, Hafiane A, Bobelin L, Canals R. Image-based crop disease detection with federated learning. Sci Rep. 2023;13:19220. doi: 10.1038/s41598-023-46218-5. DOI: https://doi.org/10.21203/rs.3.rs-3135700/v1
42. Hari P, Singh MP, Singh AK. An improved federated deep learning for plant leaf disease detection. Multimed Tools Appl. 2024;83:83471–83491. doi: 10.1007/s11042-024-18867-9. DOI: https://doi.org/10.1007/s11042-024-18867-9
43. Merenda M, Porcaro C, Iero D. Edge Machine Learning for AI-Enabled IoT Devices: A Review. Sensors. 2020;20:2533. doi: 10.3390/s20092533. DOI: https://doi.org/10.3390/s20092533
44. Ashurov AY, et al. Enhancing plant disease detection through deep learning: a Depthwise CNN with squeeze and excitation integration and residual skip connections. Front Plant Sci. 2025;15:1505857. doi: 10.3389/fpls.2024.1505857. DOI: https://doi.org/10.3389/fpls.2024.1505857
45. Anwar S, Soomro SR, Baloch SK, Patoli AA, Kolachi AR. Performance Analysis of Deep Transfer Learning Models for the Automated Detection of Cotton Plant Diseases. Eng Technol Appl Sci Res. 2023;13:11561–11567. doi: 10.48084/etasr.6187. DOI: https://doi.org/10.48084/etasr.6187
46. Rajasekar V, Venu K, Jena SR, Varthini RJ, Ishwarya S. Detection of Cotton Plant Diseases Using Deep Transfer Learning. J Mobile Multimed. 2022;18:307–324. doi: 10.13052/jmm1550-4646.1828. DOI: https://doi.org/10.13052/jmm1550-4646.1828
47. Ryo M, Schiller J, Stiller S, et al. Deep learning for sustainable agriculture needs ecology and human involvement. J Sustain Agric Environ. 2023;2:40–44. doi: 10.1002/sae2.12036. DOI: https://doi.org/10.1002/sae2.12036
48. Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D. Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization. Int J Comput Vis. 2020;128:336–359. doi: 10.1007/s11263-019-01228-7. DOI: https://doi.org/10.1007/s11263-019-01228-7
49. Lan Y, Guo Y, Chen Q, Lin S, Chen Y, Deng X. Visual question answering model for fruit tree disease decision-making based on multimodal deep learning. Front Plant Sci. 2023;13:1064399. doi: 10.3389/fpls.2022.1064399. DOI: https://doi.org/10.3389/fpls.2022.1064399
50. Lu Y, et al. Application of Multimodal Transformer Model in Intelligent Agricultural Disease Detection and Question-Answering Systems. Plants. 2024;13:972. doi: 10.3390/plants13070972. DOI: https://doi.org/10.3390/plants13070972
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Ramish Saleem, Arqam Bashir, Muhammad Arham, Ahmed Shahzad, Mehran Hanif (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
