Please login first
Applying Machine Learning for Sustainable Farm Management: Integrating Crop Recommendations and Disease Identification into Process Control and Monitoring
, *
1  School of Engineering and Technology , Department of Computer Science and engineering, GIET University, Gunupur, Odisha, India
Academic Editor: Wen-Jer Chang

Abstract:

Context: The integration of machine learning into sustainable agriculture holds immense potential for addressing the challenges faced by modern farming practices. By combining techniques for crop recommendations and disease identification, farmers can make informed decisions that not only enhance productivity but also minimize the impact of crop diseases on agricultural output.

Objective: This article aims to present a unified approach for recommending crop systems and identifying plant diseases in agricultural settings. By leveraging machine learning algorithms, the objective is to provide farmers with accurate recommendations for optimal crop selection and the timely identification of plant diseases.

Materials/Methods: Data collection for this study was facilitated using IoT sensors, including NPK sensors, DT11 sensors, and other environmental sensors, providing essential information on soil nutrients, temperature, humidity, and other environmental factors crucial for crop selection. Novel machine learning and deep learning algorithms were employed to recommend suitable crops and identify relevant plant diseases. For disease identification, real-time data from IoT sensors and high-resolution images captured by cameras were utilized. State-of-the-art convolutional neural networks (CNNs), including VGG16, ResNet50, and EfficientNetV2, were employed to accurately classify plant diseases based on visual cues such as leaf color and texture.

Results: Experimental results demonstrate that the efficacy of the proposed recommendation system utilizing CNN achieved an accuracy of 99.98%. The disease identification system, CNN, achieved a commendable accuracy of 96.06%. Subsequently, it was further deployed on cloud infrastructure, ensuring scalability and accessibility. Performance metrics such as accuracy, precision, recall, F1 Score, and AUC-ROC were used to evaluate the models' performance, with CNN exhibiting an accuracy of around 99.98%.

Conclusion: This research contributes to the advancement of sustainable farm management practices. The adoption of machine learning techniques empowers farmers to make data-driven decisions, optimize crop selection, and mitigate the impact of crop diseases, ultimately leading to improved agricultural productivity and sustainability.

Keywords: crop recommendation; machine learning ,deep learning , cloud infrastructure.
Top