Design and development of a vision-based measurement system for root growth monitoring in olives trees

Ongoing climatic changes are reshaping the environmental and edaphic conditions of agricultural soils, demanding new tools to support resilient and sustainable crop management. In this context, we present the design and validation of a low-impact, vision-based monitoring system for in-situ observation of root growth in olive trees. The system integrates a polycarbonate underground structure equipped with four fixed Arducam NOIR 8 MP cameras and red LED illumination, optimized for low-light imaging at a resolution of approximately 0.2 mm/pixel. A Raspberry Pi, managed by an Arduino with RTC module, autonomously handles image acquisition and cloud-based data transfer. Two experimental campaigns were conducted at the University of Tuscia (Viterbo, Italy). The first (June-October 2024) aimed at testing system robustness under field conditions. The system shows reliable imaging performance, but exposed limitations related to network connectivity and power management. Specifically, excessive energy consumption by the Raspberry Pi led to battery depletion in low irradiance periods. A second ongoing campaign (from May 2025) involves four systems monitoring three olive trees over a growing season, focusing on cross-plant root detection reliability. Results from the first deployment demonstrate the system's potential for non-invasive, long-term monitoring of fine root dynamics. High-quality images were obtained, and root elongation was visible across sequential captures. Planned improvements include enhanced energy efficiency and the integration of automated root segmentation algorithms. This work supports the development of scalable instrumentation for real-time phenotyping and precision agriculture under climate stress.