Cloud-Based Processing on Drone Imaging for Precision Agriculture

The agricultural industry plays a crucial role in society and consistently invests large quantities of money. This is because the process of growing crops involves varied processes that are often time-consuming. In most countries, manual labor is still a key component of agriculture. However, with current technologies, many processes in precision agriculture can be automated or improved. Among these is traditional plant health inspection, which is often time-consuming and inefficient. The research gap this paper is trying to handle is about the difficult task of continuously monitoring crops. This paper aimed to address this challenge by proposing an innovative solution for assessing plant health, which applied cloud-based processing for aerial images. It was inspired by existing drone projects, as well as the Internet of Things technology. It used data captured from drone sensors and processed it automatically in the cloud. The software architecture used to achieve this was guided by the Attribute-Driven Design (ADD) methodology. According to the findings of the study, 430 ms was the average response time for plant state classification. Moreover, the cloud architecture was capable of sending an alarm if an unusual state was reached. Finally, a colored map was created to enable better visualization.