English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Segmentation of structural parts of 3D models of plants is an important stepfor plant phenotyping, especially for monitoring architectural andmorphological traits. Current state-of-the art approaches rely on hand-crafted3D local features for modeling geometric variations in plant structures. Whilerecent advancements in deep learning on point clouds have the potential ofextracting relevant local and global characteristics, the scarcity of labeled3D plant data impedes the exploration of this potential. We adapted six recentpoint-based deep learning architectures (PointNet, PointNet++, DGCNN, PointCNN,ShellNet, RIConv) for segmentation of structural parts of rosebush models. Wegenerated 3D synthetic rosebush models to provide adequate amount of labeleddata for modification and pre-training of these architectures. To evaluatetheir performance on real rosebush plants, we used the ROSE-X data set of fullyannotated point cloud models. We provided experiments with and without theincorporation of synthetic data to demonstrate the potential of point-baseddeep learning techniques even with limited labeled data of real plants. Theexperimental results show that PointNet++ produces the highest segmentationaccuracy among the six point-based deep learning methods. The advantage ofPointNet++ is that it provides a flexibility in the scales of the hierarchicalorganization of the point cloud data. Pre-training with synthetic 3D modelsboosted the performance of all architectures, except for PointNet.

Segmentation of structural parts of rosebush plants with 3D point-based deep learning methods