Detect Plant Disease and Recommend Fertilizer and Supplement Using CNN & Mobile Net Algorithm

Abstract

Agriculture plays a crucial role in the Indian economy. Early detection of plant diseases is very much essential to prevent crop loss and further spread of diseases. Most plants such as apple, tomato, cherry, grapes show visible symptoms of the disease on the leaf. These visible patterns can be identified to correctly predict the disease and take early actions to prevent it. This can be overcome by the use of machine learning and deep learning algorithms. Hence, we are proposing a method that which is detecting the disease of a tomato plant from their leaf images. Here the process is performed with the deep learning algorithms Convolutional Neural Network (CNN), and MobileNet which is a one of the transfer learning method of CNN. Once after training the dataset with the algorithms, the accuracy of algorithms is compared and the images are classified. And the precautions are also provided for the classified plant.

Introduction

Agriculture is a major sector in India, employing more than 70% of the population, and protecting crops from diseases is essential for maintaining productivity and food security. Traditional plant disease detection methods rely on manual inspection by experts, which is time-consuming, expensive, and difficult to implement on large farms. To overcome these limitations, machine learning, deep learning, and image processing techniques are being used for automated plant disease identification.

The proposed system focuses on detecting plant diseases from leaf images by analyzing visual features such as leaf color, damage level, texture, shape, and structure. Image processing techniques extract useful information from leaf images, while machine learning algorithms use this information to classify diseases. Modern technologies such as Convolutional Neural Networks (CNN) and MobileNet improve disease detection accuracy while reducing computational requirements.

Existing disease detection approaches include laboratory-based methods such as PCR, chromatography, spectroscopy, thermography, and hyperspectral analysis. Although these methods provide accurate results, they are expensive, complex, and time-consuming. Earlier computer vision-based methods used handcrafted features and classifiers such as Support Vector Machine (SVM), Artificial Neural Networks (ANN), Decision Trees, and Random Forests. However, these systems often face challenges related to complex backgrounds, lighting variations, and multiple disease categories.

The proposed intelligent plant disease diagnosis system combines CNN and MobileNet deep learning architectures. The uploaded leaf image is first preprocessed to improve quality and standardize size. CNN extracts important visual features from diseased regions, while MobileNet performs efficient classification with lower computational cost. After identifying the disease, a recommendation module provides suitable fertilizers, micronutrients, pesticides, organic treatments, and preventive agricultural practices. The lightweight design of MobileNet allows deployment on smartphones, enabling farmers to diagnose diseases directly in the field.

The system uses a dataset containing approximately 54,305 leaf images belonging to 38 different plant disease classes, including crops such as apples, grapes, tomatoes, potatoes, corn, and strawberries. The dataset is divided into 70% training data and 30% testing data. The methodology includes dataset creation, image preprocessing, model training using CNN and MobileNet, disease classification, and accuracy evaluation. Users can upload leaf images, view prediction results, check model accuracy, and receive recommended precautions.

The CNN model performs disease classification through four main stages: convolution, pooling, flattening, and fully connected layers. Convolution extracts image features, pooling reduces dimensional complexity, flattening converts features into a suitable format, and fully connected layers perform final classification. MobileNet provides a lightweight alternative by maintaining high accuracy with reduced computational requirements, making it suitable for mobile-based agricultural applications.

Conclusion

In this project we have successfully implemented the machine learning and deep learning algorithms to build a model that which can predict the plant leaf images, either the leaf is effected with which type of disease or not along with organic, inorganic fertilizer and supplement. Here we mainly used CNN, MobileNet, for the process of training image dataset. Once after training, we have checked the classified results for the provided input image and then the precautions are provided for the diseased leaf.

References

[1] Rothe, P.R. and Kshirsagar, R.V., \"Identifying Cotton Leaf Diseases Using Pattern Recognition Techniques\", International Conference on Pervasive Computing (ICPC), 2015. [2] Aakanksha Rastogi, Ritika Arora and Shanu Sharma, “Detection and grading of leaf diseases using computer vision techniques and fuzzy logic,” 2nd International Conference on Signal Processing and Integration Networks (SPIN) 2015. [3] Godliver Owomugisha, John A. Quinn, Ernest Mwebaze und James Lwasa, ``Automated Vision-Based Diagnosis of Banana Bacterial Welt Disease and Black Sigatoka Disease“, Proceedings of the First International Conference on Mobile ICT in Africa, 2014. [4] uan Tian, Chunjiang Zhao, Shenglian Lu und Xinyu Guo, \"SVM-based Multiple Classifier System for Recognition of Wheat Leaf Diseases\", CDC\'2010 Proceedings, 20.-22. November 2010. [5] S. Yun, W. Xianfeng, Z. Shanwen und Z. Chuanlei (2015), \"Pnn-based detection of plant diseases using leaf image characteristics and meteorological data,\" International Journal of Agricultural and Biological Engineering, vol. 8, No.4, p. 60. [6] J. G. A. Barbedo (2013), \"Digital Image Processing Techniques for Detection, Quantification and Classification of Plant Diseases,\" Springer Plus, vol. 2, No. 660, pp. 1-12. [7] Caglayan, A., Guclu, O., and Can, AB. (September 2013).\"Methods for Recognizing Plants Based on Shape and Color Features of Leaf Images\" International Conference on Image Analysis and Processing (pp. 161-170). Heidelberg and Berlin:jumper. [8] X. Zheng, Z. Wang, A. Islam, I. Chan and S.Li, 2014d. \"Regression Forest for Direct Estimation of Cardiac Biventricular Volumes.\" MICCAI 2014, Medical Image Computing and Computer-Assisted Intervention. Accepted. [9] Wang P, Chen K, Yao L, Hu B, Wu X, Zhang J, et al. [10] Monica Jhuria, Ashwini Kumar, Rushikesh Borse “Image Processing for Smart Farming: Detection of Disease and Fruit Grading” Proceeding of the 2013 IEEE Second International Conference on Image Processing. [11] Sudhir Rao Rupanagudi, Ranjani B.S., Prathik Nagaraj, Varsha G. Bhat “A Cost Effective Tomato Maturity Grading System using Image Processing for 974 2015 International Conference on Green Computing and Internet of Things (ICGCIoT) Farmers” International Conference on Contemporary Computing and Information, 2014. [12] Dhaka, V. S., Meena, S. V., Rani, G., Sinwar, D., Kavita, K., Ijaz, M. F., & Wo?niak, M. (2021). A Survey of Deep Convolutional Neural Networks Applied for Prediction of Plant Leaf Diseases. Sensors, 21(14), 4749. https://doi.org/10.3390/s21144749 Cited by: 503 [13] Mohanty, S. P., Hughes, D. P., & Salathé, M. (2016). Using Deep Learning for Image-Based Plant Disease Detection. Frontiers in Plant Science, 7, 1419. https://doi.org/10.3389/fpls.2016.01419 Cited by: 6776. [14] Shafay, M. (2024). Recent advances in plant disease detection: challenges and opportunities. Plant Methods / PMC. Cited by: 92 [15] Sheneamer, A. (2024). Early detection of plant leaf diseases using stacking hybrid learning. PLOS ONE, 19(11), e0313607. https://doi.org/10.1371/journal.pone.0313607 Cited by: 14

Copyright

Copyright © 2026 Onumu Lokesh Kumar, Tulasi Miriyala. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.