Vision-based Xylem Wetness Classification in Stem Water Potential Determination

TL;DR

This work focused on automating stem detection and xylem wetness classification using the Scholander Pressure Chamber, a widely used but demanding method for SWP measurement, to refine stem detection and develop computer-vision-based methods to better classify water emergence at the xylem.

Abstract

Water is often overused in irrigation, making efficient management of it crucial. Precision Agriculture emphasizes tools like stem water potential (SWP) analysis for better plant status determination. However, such tools often require labor-intensive in-situ sampling. Automation and machine learning can streamline this process and enhance outcomes. This work focused on automating stem detection and xylem wetness classification using the Scholander Pressure Chamber, a widely used but demanding method for SWP measurement. The aim was to refine stem detection and develop computer-vision-based methods to better classify water emergence at the xylem. To this end, we collected and manually annotated video data, applying vision- and learning-based methods for detection and classification. Additionally, we explored data augmentation and fine-tuned parameters to identify the most effective models. The identified best-performing models for stem detection and xylem wetness classification were evaluated end-to-end over 20 SWP measurements. Learning-based stem detection via YOLOv8n combined with ResNet50-based classification achieved a Top-1 accuracy of 80.98%, making it the best-performing approach for xylem wetness classification.

Figures (4)

• Figure 1: Our end-to-end framework for vision-based xylem wetness classification has three distinct phases. Data collection, and dataset generation and curation help train and assess methods for stem detection. The best-performing stem detection methods are then used together with the dataset to train and evaluate several plausible learning-based xylem wetness classification methods.
• Figure 2: (a) The data acquisition setup is based on an automated pressure chamber mucchiani2024development, and allows an operator full control of the process. (b) Examples of the three classes considered in this work. (c) Leaves have been bagged for 10 minutes before the excision from the tree. (d) The bagged leaf.
• Figure 3: Sample stem detection bounding boxes via manual annotation (cyan), H20 (yellow), H30 (magenta), H40 (blue), YOLOv5nu (red), and YOLOv8n (green). The dimensions and placement of the bounding boxes differ between the tested methods. The two selected methods (H30 and YOLO8vn) along with the manually-annotated bounding box are highlighted.
• Figure 4: Top-1 Accuracy for end-to-end xylem wetness classification.