ENHANCED DISEASE CLASSIFICATION ON SOYBEAN PLANT BY USING IOT DEVICE AND MACHINE LEARNING TECHNIQUES

Abstract

Abstract Recently, there has been a substantial accumulation of data related to predicting and preventing soybean infections. These approaches vary in their methodologies and the extent of their development and application. However, their common objective is to enhance crop and product management. This study conducts assessments of various models for disease detection and pod counting in soybean plants. Specifically, two distinct models, CNN and EfficientDet B0, have been created for identifying healthy and diseased leaves while accurately quantifying pod numbers. TensorFlow, a versatile numerical computation tool, was employed in this research. This tool finds practical use in controlled farm environments, promptly detecting disease signs on plant leaves as they appear.