GPU-Powered Crop Quality Forecasting: Integrating CUDA-YOLO Detection with RF-XGBoost Classification
Abstract
Smart agriculture combines drones, artificial intelligence (AI), Internet of Things (IoT), and data analytics to solve major problems in the field, including increased food prices, water shortages, climate change, and labor shortages. IoT sensors can be used to track the health of crops in real-time, whereas edge computing can be used to make quick decisions on farms. The AI models can be used to predict diseases of plants, pest attacks, and even the optimal time to plant a particular plant, which will enhance the efficiency of resources and environmental sustainability. The given paper draws a comparison between deep learning and traditional machine learning in terms of crop quality prediction and suggests a fast and real-time CUDA-based hybrid model of RF-XGBoost and YOLO. The result of the experiments reveals that the suggested method has a precision of approximately 96 percent, which is higher than the accuracy of SVM (88 percent) and Decision Tree (91 percent) in precision, speed and scale and can support data-driven and sustainable precision agriculture.