Authors:
S. Sujitha, A. Singaravelan, M. Arunadevi Thirumalraj, Thirumalraj Karthikeyan, Dilli Kasi Rao Kotha

Addresses:
Department of Electrical and Electronics Engineering, New Horizon College of Engineering, Bengaluru, Karnataka, India. Department of Electrical and Electronics Engineering, Nitte Meenakshi Institute of Technology, Bengaluru, Karnataka, India. Department of Computer Science and Engineering, Karunya Institute of Technology and Science, Coimbatore, Tamil Nadu, India. Department of Computer Science and Business Management, Saranathan College of Engineering, Tiruchirappalli, Tamil Nadu, India. Department of Artificial Intelligence, Trichy Research Labs, Quest Technologies, Tiruchirappalli, Tamil Nadu, India. Faculty of Engineering, Environment and Computing, Coventry University, Coventry, England, United Kingdom.

Abstract:

Rice is needed daily. This food contains carbs, which the body requires for energy and growth. Economically beneficial rice. Infected rice plants waste field crop and reduce yield if not monitored. Field monitoring is needed to detect rice plant illnesses. Recent agricultural surveillance uses drones with cameras and GPS. It's quick, independent data collection over a vast area. This study proposes a drone-based (IoT) building that detects and classifies rice diseases using real-time data and image processing to boost rice productivity. The method uses a GPS sensor to map ill rice plants in rice fields in real time. This study suggests using the COA to determine the optimal network hyperparameter values for autonomous rice image categorisation. The Swin-Transformer links non-overlapping windows from the previous layer by changing window partitions, thereby capturing multi-scale properties. The improved approach detects six categories, including healthy, leaf blast, and bacterial leaf blight. Experimental analysis used Accuracy, Precision, Recall, Specificity, and F1_score. For training 80% and testing 20% data, the Chimpanzee-based model obtained 96% accuracy, 99% specificity, 96.09% F1_score, 96.26% precision, and 96.14% recall, a 3% to 5% improvement over earlier techniques. The suggested enhancement enhanced literature findings compared to other methods using the same or similar datasets.

Keywords: Image Processing; GPS Sensor; Rice Plant Disease Classification; Internet of Things (IoT); Multi-Scale Features; Chimpanzee Optimisation Algorithm; Swin Transform; Chimpanzee Optimisation Algorithm (COA).

Received on: 13/10/2024, Revised on: 17/12/2024, Accepted on: 29/01/2025, Published on: 14/09/2025

DOI: 10.69888/FTSESS.2025.000539

FMDB Transactions on Sustainable Environmental Sciences, 2025 Vol. 2 No. 3, Pages: 120-136

• Views : 106
• Downloads : 12

Download PDF