International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 12 | Dec 2024
p-ISSN: 2395-0072
www.irjet.net
Raita Mitra for Crop, Fertilizer and Plant Disease Detection using ML Shalini B N 1 and Priya M K2 1Senior Grade Lecturer
Department of Computer Science & Engineering, Government Residential Women’s polytechnic, Shivamogga, Karnataka, India 2Senior Grade Lecturer,Department of Computer Science &Engineering,Government polytechnic,Ramanagara, Karnataka, India -------------------------------------------------------------------------***-----------------------------------------------------------------------Abstract This study presents a robust agricultural decision support system employing machine learning techniques for crop recommendation, fertilizer prescription, and disease anticipation, integrating essential agricultural indicators such as soil nutrient levels, pH, precipitation trends, and crop varieties. Through preprocessing and model training, the system utilizes Random Forest and Naive Bayes algorithms for crop and fertilizer prediction, achieving commendable accuracy rates of 99.09% and 99.2%, respectively, while disease prediction relies on the ResNet-9 model. By empowering farmers with wellinformed choices regarding crop selection, optimal fertilization techniques, and disease control, the system enhances agricultural efficiency and sustainability. This research significantly contributes to precision agriculture, promoting sustainable methodologies and safeguarding food security through enhanced crop productivity and optimized resource allocation.
Keywords—Convolutional Neural Networks, Disease Detection Deep Learning, Agricultural Productivity. Crop Recommendation & Fertilizer Recommendation, Machine Learning.
I. INTRODUCTION
revolutionizing farming through data-driven decisionmaking and innovative solutions.
This project harnesses machine learning to assist farmers in making informed decisions regarding crop selection, fertilizer application, and disease control. By analyzing factors like soil nutrients, rainfall patterns, crop types, and leaf images, it delivers personalized recommendations. Utilizing advanced algorithms such as Random Forest and Naive Bayes, the system aims for high prediction accuracy, ultimately optimizing agricultural practices for increased productivity and sustainability. Through the integration of diverse data sources, it addresses key challenges faced by farmers, aiming to enhance crop yield while minimizing resource wastage and environmental impact.
II LITERATURE REVIEW SamyakShrimali et al. achieved an accuracy of 95.7% and an F1 score of 96.1% in detecting crop diseases using the MobileNetV2 model architecture and image filters [1]. Leninisha Shanmugam et.al. proposed a methodology for the automatic detection of plant diseases using remote sensing images [2]. The model described in the research papers makes use of the PlantVillagpreventpe dataset, which contains a substantial number of images for training and testing. Specifically, the dataset used in the research includes 11993 images with 11 different classes of plant diseases [3].54,305 photos of 38 different plant disease classes [4], and a dataset providing good variations in color, orientation, and size of leaves for plant disease identification [5]. These diverse datasets enable the models to learn and generalize effectively, resulting in high accuracy rates ranging from 97.73% to 99.80% in disease classification and identification tasks. The utilization of such extensive datasets is crucial for training robust models capable of accurately identifying various plant diseases and pests in agricultural settings. The crop recommendation models discussed in the provided contexts utilize machine learning algorithms like Random Forest, KNN, Decision Tree, and others to suggest suitable crops based on factors such as soil nutrient values (N, P, K),
The developed system provides a user-friendly interface for easy access and interpretation of recommendations, empowering farmers with actionable insights and contributing to the advancement of precision agriculture. Combining cutting-edge technology with practical solutions to real-world agricultural problems, the project bridges the gap between traditional farming methods and modern data-driven approaches. Collaboration with agricultural experts ensures relevance and applicability to the farming community, with an emphasis on simplicity and usability to cater to the diverse needs of farmers worldwide. Driven by a commitment to improving food security and promoting sustainable farming practices, the project envisions a future where technology serves as a powerful tool for agricultural development,
© 2024, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 681