Real-Time Crop Growth Tracking and Disease Detection using Machine Learning

• Published in: Proceedings (International Confernce on Computational Intelligence and Communication Networks) pp. 457 - 461
• Main Authors: Babu, A Rajendra, Mohebbanaaz, Lalitha, T., Anjali, B., Sree, U. Chaithya
• Format: Conference Proceeding
• Language: English
• Published: IEEE 22.12.2024
• Subjects: Accuracy; Agriculture; Crop Recommendations; Crops; Data mining; disease detection; Feature extraction; Machine learning; Pesticides; Random forests; Real-time crop growth; Real-time systems; Resilience; Soil; Support vector machines
• ISSN: 2472-7555
• DOI: 10.1109/CICN63059.2024.10847369

Real-time tracking and analysis of crops are applied to monitor and assess the key variables that affect crop growth and development, such as the weather, soil, and plant health indicators. Through this system, decisions regarding the use of resources can be made more effectively, ensuring increased productivity. It fosters sustainable agriculture and enhanced resilience to changing environmental conditions, thus ensuring long-term food security. SVM and Random Forest, which are kinds of learning machines, play an essential role in this kind of process. Crop Recommendation data set is extracted from Kaggle website [1]. The overall accuracy obtained for Random Forest is 98.54 % and accuracy obtained for SVM is 99.16 %. The generalization power of its prediction made by SVM is more acceptable since it's accuracy may be highly sensitive to the extracted features. Also, computational cost may make it slow for big datasets.