Year: 2026
Type: article
Abstract: This paper presents a rail-guided IoT imaging system designed to overcome the difficulty of obtaining consistent, leaf-level citrus images for long-term agricultural monitoring by combining deterministic electromechanical positioning, embedded control, edge computing, and cloud-based data management. It incorporates a cable-suspended linear actuation mechanism together with two inclined cameras, motion control based on ESP32 microcontrollers, a Raspberry Pi 5 for processing, and AWS S3 for cloud services to automate the processes of image acquisition, synchronization, storage, and remote visualization. By limiting image capture to a fixed and repeatable path, the platform reduces the variation in viewpoint that is typically found with aerial and ground-mobile robots, thus allowing the systematic production of consistent image datasets for use in plant phenotyping and pest monitoring. The system was implemented and tested in a citrus orchard under real-world operating conditions. Experimental results demonstrate the system's autonomous and stable operation, reliable image acquisition, continuous synchronization with the cloud and repeatable monitoring over long periods of time with a low-cost hardware design suitable for research-scale setups. The platform provides a reproducible embedded imaging setup, which improves the consistency of agricultural image acquisition and enables the development and evaluation of computer vision algorithms for precision agriculture.
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
Open Access status: diamond
OpenAlex ID: W7225903091
FWCI:
Cites:
Cited by:
Related to: