Potato Plant Disease Detection

Abstract

Potatoes are widely grown worldwide and play a significant role in many cuisines, especially in India, where they are among the top crops cultivated. Potato plants often get sick, which can make them not very good to eat and make fewer potatoes grow. For maximum production it is necessary to maintain crop health by identifying various kinds of diseases. Similar to approaches used to identify illnesses in tomato leaves, advancements in deep learning and machine learning techniques appear potential for automating the identification and treatment of potato issues. These approaches involve steps like preparing datasets, processing images, extracting features, and training models. Various algorithms and models like VGG-16, U-Net, Inception-v3, PLeaD-Net, and PLDPNet have been utilized, demonstrating high accuracies in recognizing disease patterns in plant images. This improvement in detecting potato plant diseases helps farmers take better care of their crops and find ways to make them healthier and produce more potatoes.

Introduction

Potato farming is crucial in countries like India for food security and economic growth but faces significant threats from diseases that affect crop health and yield. Traditional disease detection methods are slow and rely on expert judgment, making them unsuitable for large-scale farming. To address this, recent research focuses on using advanced technologies like deep learning, machine learning, and image processing for fast, accurate, and cost-effective disease detection in potato plants.

Several deep learning models and architectures have been explored and developed for this purpose:

• CNN-based models use convolutional layers to extract features and classify potato leaf diseases with high accuracy.

• VGG-16 and ResNet are popular deep learning algorithms applied for disease classification, achieving accuracies often above 98%.

• Hybrid models like PLDPNet combine segmentation (using U-Net) and classification (using VGG19 and Vision Transformer) for precise disease prediction.

• Lightweight models like PLeaD-Net are designed for early disease detection on resource-constrained devices, incorporating attention mechanisms for enhanced performance.

• Transfer learning approaches (Inception-v3, ResNet50) leverage pre-trained networks to improve classification accuracy on potato leaf images.

• Studies show that integrating image preprocessing, segmentation, feature extraction, and classification techniques leads to efficient and accurate disease detection.

The ultimate goal is to enable timely and precise identification of diseases, reducing pesticide use and protecting farmers’ livelihoods. These AI-driven techniques promise scalable, automated solutions suitable for modern agriculture, especially benefiting smallholder farmers.

Conclusion

Our study focused on making a computer program to spot diseases in potato plants using a type of computer system called a convolutional neural network (CNN). We trained this program with lots of pictures of sick tomato plants to see how well it could identify different diseases. We checked how good it was by looking at numbers like F1-score, recall, accuracy, and precision. Our output showed proved the program was really good at finding potato plant diseases, with an accuracy of about 98% on the test pictures. It could tell apart different types of diseases quite well. We also looked at other similar studies and found that many different computer models, like CNNs, VGG-16, PLDP-Net, Plead-Net, and InceptionV3, did a good job too, with accuracies ranging from 95% to 98%. These models could spot diseases like blight in early stages and late stages. When we compared our program to others like InceptionV3, VGG-16, PLDP-Net, and Plead-Net, we saw that ours was simpler and used less computer power, which makes it more practical for finding potato plant diseases.

References

[1] Aniket Gahiware!, Nikhil Bawake, Bhushan Dhage, Lokesh Ahir, Prof. V. K. Abang “Potato Plant Disease Detection using Deep Learning” International Journal of Advanced Research in Science, Communication and Technology,Volume 2, Issue 2, May 2022. [2] Girish B G!, Narendra N2, Nitish S°, Manoj M D4, R Virupakshagouda Patil’ “Advanced Machine Learning Technique to Detect Disease in Potato”. [3] Arshad, Fizzah, Muhammad Mateen, Shaukat Hayat, Maryam Wardah, Zaid Al-Huda, Yeong Hyeon Gu, and Mugahed A. Al-antari. \"PLDPNet: End-to-end hybrid deep learning framework for potato leaf disease prediction.\" Alexandria Engineering Journal 78 (2023): 406-418. [4] Eraj, Sama Uddin, and Mohammed Nazim Uddin. \"Early stage Potato Disease Classification by analyzing Potato Plants using CNN.\" In 2023 International Conference on Next-Generation Computing, IoT and Machine Learning (NCIM), pp. 1-6. IEEE, 2023. [5] Gardie, B., S. Asemie, K. Azezew, and Z. Solomon. \"Potato Plant Leaf Diseases Identification Using Transfer Learning.\" Indian Journal of Science and Technology 15, no. 4 (2022): 158-165. [6] Radhika, P., P. Tejeswara Murthy, G. Pavaneeshwar Reddy, V. Durga Vara Prasad, T. Harshitha, and N. B. V. Bharath. \"Potato leaf disease detection using convolutional neural networks.\" (2023). [7] Sholihati, Rizqi Amaliatus, Indra Adji Sulistijono, Anhar Risnumawan, and Eny Kusumawati. \"Potato leaf disease classification using deep learning approach.\" In 2020 international electronics symposium (IES), pp. 392-397. IEEE, 2020. [8] Singh, Gulbir, and Kuldeep Kumar Yogi. \"Comparison of RSNET model with existing models for potato leaf disease detection.\" Biocatalysis and Agricultural Biotechnology 50 (2023): 102726. [9] Asif, Md Khalid Rayhan, Md Asfaqur Rahman, and Most Hasna Hena. \"CNN based disease detection approach on potato leaves.\" In 2020 3rd International conference on intelligent sustainable systems (ICISS), pp. 428-432. IEEE, 2020.

Copyright

Copyright © 2025 Dr. Avinash Bhute, Aniket Gavkare, Pratik Nagare, Vineet Morankar, Moin Shaikh. 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.