Autonomous Bell Pepper Disease Detection Robot Using Artificial Intelligence and Deep Image Processing: A Case Study of Quanta Synergy Co., Ltd

Abstract

This research presents an autonomous robot for detecting bell pepper diseases using Artificial Intelligence (AI) and Deep Learning for plant image analysis and disease classification. The robot navigates autonomously in greenhouses using the strategy of Ko et al. (2015) and employs Deep Convolutional Neural Networks (CNNs) based on Picón et al. (2019) for classifying four diseases white spot, anthracnose, mosaic virus, and yellow leaf curl. The robot is equipped with a Logitech B525 webcam, LiDAR (Robosense RSLiDAR-16), and other sensors for disease and obstacle detection. The system achieves high accuracy in disease classification 90% for white spot, 100% for anthracnose, 100% for mosaic virus, 90% for yellow leaf curl, and 97.5% for healthy leaves. The robot is powered by an 882 Wh battery and has an average power consumption rate of 22.5 W. The research results demonstrate that the robot helps reduce farmers' workload, shortens disease inspection time, improves disease control efficiency, reduces chemical usage, and promotes smart agriculture. The results confirm that the robot can be effectively deployed in real agricultural environments and can significantly enhance the accuracy and efficiency of smart farming systems.