Abstract

Leaf vein network geometry can predict levels of resource transport, defence and mechanical support that operate at different spatial scales. However, it is challenging to quantify network architecture across scales due to the difficulties both in segmenting networks from images and in extracting multiscale statistics from subsequent network graph representations. Here we developed deep learning algorithms using convolutional neural networks (CNNs) to automatically segment leaf vein networks. Thirty-eight CNNs were trained on subsets of manually defined ground-truth regions from >700 leaves representing 50 southeast Asian plant families. Ensembles of six independently trained CNNs were used to segment networks from larger leaf regions (c. 100mm2 ). Segmented networks were analysed using hierarchical loop decomposition to extract a range of statistics describing scale transitions in vein and areole geometry. The CNN approach gave a precision-recall harmonic mean of 94.5%±6%, outperforming other current network extraction methods, and accurately described the widths, angles and connectivity of veins. Multiscale statistics then enabled the identification of previously undescribed variation in network architecture across species. We provide a LeafVeinCNN software package to enable multiscale quantification of leaf vein networks, facilitating the comparison across species and the exploration of the functional significance of different leaf vein architectures.

Highlights

  • Plant leaves are structured by a vein network that ranges in geometry from dendritic fans with few branches and no loops (e.g. Ginkgo biloba) to hierarchical forms with many loops (e.g. Acer saccharum) (Trivett & Pigg, 1996; Roth-Nebelsick et al, 2001)

  • To illustrate the accuracy of the ensemble convolutional neural networks (CNNs) approach, we show weighted networks for six species (Artocarpus odoratissimus, Dryobalanops lanceolata, Lophopetalum javanicum, Macaranga pearsonii, Pentace laxiflora, and Terminalia citrina) covering a range of different vein architectures from dense loops to open trees (Fig. 4)

  • The ensemble CNN applied to this ROI provided a smooth, high-contrast probability map of vein identity (Fig. 5b) that was thresholded to give a full-width binary (FWB) image (Fig. 5b0) and compared with the GT using P-R analysis (Fig. 5l)

Read more

Summary

Introduction

Plant leaves are structured by a vein network that ranges in geometry from dendritic fans with few branches and no loops (e.g. Ginkgo biloba) to hierarchical forms with many loops (e.g. Acer saccharum) (Trivett & Pigg, 1996; Roth-Nebelsick et al, 2001). The vein density (vein length per unit area) at each order, or across all orders, is a key statistic (Uhl & Mosbrugger, 1999; Sack & Scoffoni, 2013). Measurement of the distribution of vein radii is required to determine the vein order (Price et al, 2012), which is itself challenging. Vein orders can be extracted using hierarchical vein classification methods (Gan et al, 2019). If vein radii can be extracted reliably, they can be used to

Methods
Results
Discussion
Conclusion

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.