Design and Development of a Neural Network-Based End-Effector for Disease Detection in Plants with a 7 DOF Robot Integration

Abstract

Agriculture and robotics have managed to integrate, using artificial intelligence in plant tracking and pest detection, as well as robotic arms and remote-controlled robots for harvesting, significantly reducing the human workload. Robots are typically designed to perform specific tasks, making their adaptability very difficult to integrate into agriculture due to the constant changes of plants, such as plant growth. As a result of the general and current functions in agro-robotics, continuous monitoring with deep learning is aimed at knowing the condition of the plants and that the mobility of the robots does not impede the plant’s growth while being monitored for it to adapt to the monitoring environment. Deep learning and a robotic arm are used for real-time plant monitoring. Image database will be used for training, accuracy, recall and F1 indicators are used to evaluate the network. The robot has kinematics that allows it to change its size to monitor plant health and growth. Early detection of plant leaf anomalies and diseases using a deep learning system. Integrating the various systems, provided an intelligent and effective solution for detecting anomalies and diseases in the leaves of plants subjected to intelligent robotic monitoring.

Code and data availability

公開論文であることは確認できましたが、現在の公式API・許可済み取得経路では本文を自動取得できませんでした。

No evidence-backed public reproduction asset is currently recorded.