The \"Potato Leaf Disease Detection\" project aims to solve the problem of accurately identifying and diagnosing diseases in potato plants, specifically focusing on late blight and early blight. Late blight is caused by the microorganism Phytophthora infestans, and early blight is caused by the fungus Alternaria solani,. These diseases can significantly impact potato yields, resulting in substantial economic losses for farmers. By developing a mobile application, we provide farmers with a tool to quickly and accurately diagnose these diseases, enabling timely and appropriate interventions to mitigate crop damage. The core functionality of the application involves capturing images of potato plant leaves and analyzing them to predict the presence of early blight or late blight. The application uses advanced machine learning (ML) techniques, mainly Convolutional Neural Networks (CNNs), to achieve high accuracy in disease detection. Users simply need to take a picture of a potato leaf, and the application will provide a diagnosis with confidence percentage, enabling farmers to make informed decisions about disease management. This project not only aims to provide immediate benefits to farmers by identifying specific diseases but also sets the foundation for future enhancements. By expanding the application to detect a wider range of plant diseases and incorporating additional features such as pest management tips, weather forecasts, and soil health analysis, the project envisions creating a comprehensive farming support tool. The ultimate goal is to improve crop productivity, reduce losses, and enhance the economic well-being of farmers through the effective use of technology.
Introduction
Potato, a crucial global crop, suffers significant yield losses due to diseases like late blight and early blight. Early and accurate disease detection is vital to reduce economic impact, but traditional methods are slow, costly, and require expert knowledge, making them impractical for many farmers, especially in remote areas. This research develops an intelligent, automated potato leaf disease detection system using deep learning and image segmentation. The system leverages Convolutional Neural Networks (CNNs) trained on a Plant Village dataset, achieving over 99% accuracy. Integrated into a user-friendly web application, it enables farmers without technical expertise to diagnose diseases early and accurately.
Bangladesh, a top potato producer, faces challenges with diseases affecting production and export. Effective disease management through AI-powered detection can improve yield and sustainability.
The literature review highlights various machine learning and image processing techniques previously used for potato disease detection, including light wavelength analysis, segmentation, transfer learning, and UAV-based data collection. These studies confirm the effectiveness of CNNs and transfer learning in improving disease classification accuracy.
The proposed approach involves preprocessing (resizing, noise filtering, color normalization), segmentation to isolate leaf regions and diseased spots, feature extraction (color, texture, structure), and CNN model training for classification into healthy or diseased categories (early blight, late blight, bacterial wilt). The model undergoes rigorous testing and validation, ensuring reliable performance. Final results can be displayed via web or mobile applications, supporting timely decision-making by farmers.
Conclusion
This project focuses on detecting potato diseases using CNN, as it has proven to be the most effective method, achieving 99% validation accuracy. With a large dataset, we worked hard to ensure accuracy and believe this project can greatly benefit the agricultural sector. Many farmers in Bangladeshi villages are unaware of disease detection methods, leading to crop destruction by insects and significant losses. Our work aims to change this by providing a simple and accessible solution. In the future, we plan to develop an Android app to detect diseases in various crops and offer proper solutions. Expanding our dataset will further enhance accuracy. By creating this system, farmers will receive instant advice and take quick action to protect their crops, improving agricultural productivity and reducing losses.
References
[1] Shukla, P. K. \"Early Detection of Potato Leaf Diseases using CNN with Web Application.\" [Online]. Available: https://www.researchgate.net/publication/380829683_Potato_Leaf_Disease_Detection_By_Deep_Learning_A_Comparative_Study. [Accessed: 24-Mar-2025].
[2] Akter, S., & Sattar, A. \"Potato Leaf Diseases Detection using Deep Learning.\" [Online]. Available: https://www.researchgate.net/publication/357774350_Detection_and_Classification_of_Potato_Diseases_Using_a_New_Convolution_Neural_Network_Architecture. [Accessed: 24-Mar-2025].
[3] Ghosh, H., & Rahat, I. S. \"Potato Leaf Disease Recognition & Prediction using CNN.\" [Online]. Available: https://www .researchgate. net/publication/ 374770403_Detecting_Diseases_in_Potato_Leaves_using_Deep_Learning_and_Machine_Learning_Approaches_A_Review. [Accessed: 24-Mar-2025].
[4] Kothari, D., & Gharat, M. \"Potato Leaf Diseases Detection using Deep Learning.\" [Online]. Available: https://www.researchgate. net/publication/380829683_ Potato_Leaf_Disease_Detection_By_Deep_Learning_A_Comparative_Study. [Accessed: 24-Mar-2025].
[5] Fernandez, M., Zhao, N., & Mehta, O. \"Enhancing Potato Disease Detection Using Transfer Learning.\" [Online].Available:https://www.researchgate.net/publication/357774350_Detection_and_Classification_of_Potato_Diseases_Using_a_New_Convolution_Neural_Network_Architecture. [Accessed: 24-Mar-2025].
[6] Williams, P., Zhang, Q., and Lee, R. (2025). \"Early Detection of Potato Diseases through UAVs and Artificial Intelligence.\" MDPI Journal of Remote Sensing, 15(9), 2450. Available online: https://www.mdpi.com/2072-4292/15/9/2450. Accessed on: March 24, 2025.
[7] Yuan, D., Wu, C., & Li, J. (2020). \"Detection of Potato Leaf Diseases Using a Hybrid Convolutional Neural Network.\" IEEE Access, 8, 72671-72677.
[8] Dickson, M.A., & Bausch, W.C. (1997). \"Plant Recognition via Neural Network Classifier with Size and Shape Descriptors.\" Transactions of the ASAE, 1, 97-102.
[9] Bashir, S., & Sharma, N. (2012). \"Plant Disease Detection in Remote Areas Using Image Processing,\" Sep-Oct.
[10] Wojnar, L. (1999). Image Analysis: Applications in Material Engineering. CRC Press LLC.
[11] Bernardes, A.A., Rogeri, J.G., Marranghello, N., & Pereira, A.S. (2001). \"Identification of Foliar Diseases in Cotton Crops.\" Vol. 17, No. 1, March.
[12] Anami, B.S., Pujari, J.D., &Yakkundimath, R. (2011). \"Identification and Classification of Normal and Affected Agricultural/Horticultural Produce Based on Combined Colour and Texture Features.\" Vol. I, No. III, September.
[13] Viraj A. Gulhane and Ajay A. Gurjar, \"Detection of diseases on cotton leaves and its possible diagnosis\", International Journal of Image Processing (IJIP), vol. 5, no. 5, 2011.
[14] Sanjay B. Patil and Shrikant K. Bodhe, \"Leaf disease severity measurement using image processing\", IEIT Journal of Adaptive & Dynamic Computing, no. 4, pp. 25-30, 2012.
[15] Piyush Chaudhary, Anand K. Chaudhari, A. N. Cheeran and Sharda Godara, \"Color transform-based approach for disease spot detection on plant leaf\", International Journal of Computer Science and Telecommunications, vol. 3, no. 6, June 2012.
[16] Tushar H Jaware, Ravindra D Badgujar and Prashant G Patil, \"Crop disease detection using image segmentation\", World Journal of Science and Technology, vol. 2, no. 4, pp. 190-194, 2012
[17] Javed Rashid, Imran Khan, Ghulam Ali, Sultan H. Almotiri, Mohammed A. AlGhamdi and Khalid Masood, \"Multi-level deep learning model for potato leaf disease recognition\", MDPI.
[18] C. U. Kumari, S. Jeevan Prasad and G. Mounika, \"Leaf disease detection: feature extract ion with k-means clustering and classification with ann\", 2019 3rd International Conference on Computing Methodologies and Communication (ICCMC) Erode India 2019, pp. 1095-1098.
[19] Guanlin Li, Zhanhong Ma and Haiguang Wang, \"Image recognition of grape downy mildew and grape powdery mildew based on support vector machine\", IFIP Advances in Information and Communication Technology, vol. 37, pp. 151-162, 2011.
[20] Jing Chen, Qi Liu and Lingwang Gao, \"Visual tea leaf disease recognition using a convolutional neural network model\", Symmetry, vol. 11, pp. 343, 2019.