Leaf Disease Detection Using VGG16-Based CNN Model in a Web Application

Abstract

Nowadays, plant diseases (mainly leaf diseases) have seriously affected agricultural production. Early identification and control are crucial for minimizing crop loss and maximizing yield potential. Manual observation by experts is a common approach to detecting leaf disease; however, this method is typically time-consuming and prone to errors. With the emergence of machine learning and deep learning technologies, automated systems for detecting leaf diseases have become more effective and more robust. This article describes a web-based solution that automates leaf disease detection and identification using the VGG-16 deep learning model, trained on images sent by users. The VGG16 model is known for its deep CNN architecture, which has been pre-trained on the ImageNet dataset. After transfer learning was performed, it was used to learn various plant diseases from leaf images. Moreover, the web application presents actionable feedback to farmers by recommending identified supplements and treatments for the diagnosed diseases. This is achieved by associating each detected disease with a corresponding set of agricultural supplements. The app, built using the Flask framework, features a user-friendly interface that enables users to upload leaf pictures and receive disease predictions, as well as treatment suggestions. By embedding deep learning as one component in the decision-making chain of agriculture, the proposed technique aims to help farmers diagnose crops accurately and in a timely manner, ensuring that crop management is well-maintained, which will ultimately lead to sustainable agriculture. The system\'s real-time usability in the fields is a revolutionary shift in plant disease management, which will have major implications for increasing agricultural yield and crop protection.

Introduction

The global agricultural industry faces significant challenges from pests, unpredictable weather, and plant diseases, particularly those affecting leaves, which can cause substantial crop loss—around 20% of food is lost annually due to disease. Traditional disease detection relies on visual inspection, which is time-consuming, labor-intensive, and prone to errors, especially in early-stage infections.

Deep learning, particularly Convolutional Neural Networks (CNNs) like VGG16, has shown high accuracy in image classification and has been successfully applied to plant disease detection. This paper focuses on fine-tuning the VGG16 model through transfer learning to identify multiple leaf diseases efficiently. The system is integrated into a user-friendly web application that not only diagnoses diseases but also provides tailored recommendations for plant treatment, allowing farmers to take timely action.

The methodology involves using the PlantVillage dataset (61,486 images) with data augmentation techniques (flipping, rotation, noise injection, scaling, etc.) to train the VGG16 model. The system workflow includes image acquisition, preprocessing, segmentation, feature extraction, classification, and supplement recommendation.

Results show that the web application accurately identifies leaf diseases and provides actionable treatment suggestions, making it both diagnostic and prescriptive. Previous studies support VGG16’s high accuracy (91–99.2%) across various crops, demonstrating the effectiveness of transfer learning, data augmentation, and mobile/web deployment for real-time disease detection in precision agriculture.

Conclusion

The developed web application demonstrates the feasibility of using the VGG16 model for leaf disease detection and supplement recommendation. This approach can aid farmers in early disease detection and informed decision-making, contributing to sustainable agricultural practices. Image processing offers an efficient method for improving agricultural yield through plant disease detection. The primary objective of this project was to evaluate the capability of image processing tools in accurately identifying plant diseases and supporting farmers in increasing crop yields. The project successfully introduced image processing techniques for plant disease detection. Developing a standalone application will enhance the accessibility and utility of this technology for farmers. A dedicated system can be developed to distinguish between diseased and healthy plants. Future work will focus on creating a mobile application to further assist farmers and employing drones to expand the training image dataset, with the goal of improving the system\'s accuracy and generalization. The current system utilizes a feature-based approach achieved through image processing techniques. The process involves several steps: image acquisition, preprocessing, segmentation, feature extraction, and classification. The proposed system employs a convolutional neural network (CNN), which achieved an accuracy of 96.23 percent. The VGG16 model was also applied for leaf disease prediction, but it demonstrated lower accuracy than the CNN. Future work may include expanding the number of leaf and disease type classes.

References

[1] S. M. Metev and V. P. Veiko, Laser Assisted Microtechnology, 2nd ed., R. M. Osgood, Jr., Ed. Berlin, Germany: Springer-Verlag, 1998. [2] J. Breckling, Ed., The Analysis of Directional Time Series: Applications to Wind Speed and Direction, ser. Lecture Notes in Statistics. Berlin, Germany: Springer, 1989, vol. 61. [3] S. Zhang, C. Zhu, J. K. O. Sin, and P. K. T. Mok, “A novel ultrathin elevated channel low-temperature poly-Si TFT,” IEEE Electron Device Lett., vol. 20, pp. 569–571, Nov. 1999. [4] M. Wegmuller, J. P. von der Weid, P. Oberson, and N. Gisin, “High resolution fiber distributed measurements with coherent OFDR,” in Proc. ECOC’00, 2000, paper 11.3.4, p. 109. [5] R. E. Sorace, V. S. Reinhardt, and S. A. Vaughn, “High-speed digital-to-RF converter,” U.S. Patent 5 668 842, Sept. 16, 1997. [6] (2002) The IEEE website. [Online]. Available: http://www.ieee.org/ [7] M. Shell. (2002) IEEEtran homepage on CTAN. [Online]. Available: http://www.ctan.org/tex-archive/macros/latex/contrib/supported/IEEEtran/ [8] FLEXChip Signal Processor (MC68175/D), Motorola, 1996. [9] “PDCA12-70 data sheet,” Opto Speed SA, Mezzovico, Switzerland. [10] A. Karnik, “Performance of TCP congestion control with rate feedback: TCP/ABR and rate adaptive TCP/IP,” M. Eng. thesis, Indian Institute of Science, Bangalore, India, Jan. 1999. [11] J. Padhye, V. Firoiu, and D. Towsley, “A stochastic model of TCP Reno congestion avoidance and control,” Univ. of Massachusetts, Amherst, MA, CMPSCI Tech. Rep. 99-02, 1999. [12] Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specification, IEEE Std. 802.11, 1997. [13] http://dx.doi.org/10.17 (J & GOPAL, 2019), A. P., & GOPAL, G. (2019), https://data.mendeley.com/datasets/tywbtsjrjv/1.Retrieved, http://dx.doi.org/10.17632/tywbtsjrjv.1.

Copyright

Copyright © 2025 Vikramaditya Tarai, Dr. Vidya Mohanty, Bibhuprasad Subudhi, Rudra Prasad Padhy. 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.