Leafeon: Towards Accurate, Robust and Low-cost Leaf Water Content Sensing Using mmWave Radar

TL;DR

Leafeon uses mmWave radar with deep learning to accurately and robustly measure leaf water content non-invasively, outperforming existing methods and enabling drought detection in various environments.

Contribution

This paper introduces Leafeon, a novel mmWave radar-based system utilizing electronic beam steering and deep neural networks for precise, robust, and low-cost leaf water content sensing.

Findings

Achieved MAE as low as 3.17% in lab conditions.

Significantly outperformed state-of-the-art methods with up to 55.7% error reduction.

Successfully detected drought conditions in live plant experiments.

Abstract

Plant sensing plays an important role in modern smart agriculture and the farming industry. Remote radio sensing allows for monitoring essential indicators of plant health, such as leaf water content. While recent studies have shown the potential of using millimeter-wave (mmWave) radar for plant sensing, many overlook crucial factors such as leaf structure and surface roughness, which can impact the accuracy of the measurements. In this paper, we introduce Leafeon, which leverages mmWave radar to measure leaf water content non-invasively. Utilizing electronic beam steering, multiple leaf perspectives are sent to a custom deep neural network, which discerns unique reflection patterns from subtle antenna variations, ensuring accurate and robust leaf water content estimations. We implement a prototype of Leafeon using a Commercial Off-The-Shelf mmWave radar and evaluate its performance with a variety of different leaf types. Leafeon was trained in-lab using high-resolution destructive leaf measurements, achieving a Mean Absolute Error (MAE) of leaf water content as low as 3.17% for the Avocado leaf, significantly outperforming the state-of-the-art approaches with an MAE reduction of up to 55.7%. Furthermore, we conducted experiments on live plants in both indoor and glasshouse experimental farm environments (see Fig. 1). Our results showed a strong correlation between predicted leaf water content levels and drought events.