Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering

X-ray computed tomography (CT) is a powerful tool for in situ plant root system architecture (RSA) characterization. Accurate root segmentation from CT images is integral to studying RSA. Research studies on segmenting roots from CT images have been mainly limited to image processing-based approache...

Full description

Saved in:
Bibliographic Details
Main Authors: Mary E. Cassity, Paul C. Bartley, Yin Bao
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Smart Agricultural Technology
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2772375524002715
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846124936025866240
author Mary E. Cassity
Paul C. Bartley
Yin Bao
author_facet Mary E. Cassity
Paul C. Bartley
Yin Bao
author_sort Mary E. Cassity
collection DOAJ
description X-ray computed tomography (CT) is a powerful tool for in situ plant root system architecture (RSA) characterization. Accurate root segmentation from CT images is integral to studying RSA. Research studies on segmenting roots from CT images have been mainly limited to image processing-based approaches which may require parameter tuning and often lack common segmentation metrics, e.g., Dice and IoU. A recent deep learning approach utilizes a volumetric encoder-decoder network to achieve a high Dice score and IoU. However, training a volumetric model is dependent on the availability of fully annotated scans of the growing medium column, obtaining which can be time-consuming, tedious, and resource intensive. In this study, an efficient method using deep learning-based instance segmentation in conjunction with density-based spatial clustering of applications with noise (DBSCAN)-based filtering was developed and evaluated for two horticultural plant species. A pretrained Mask R-CNN model was fine-tuned on images selected along different axes of the three-dimensional scans to identify the best view selection strategy for volumetric root segmentation. DBSCAN was used to filter noise from the volumetric segmentation with an automated parameter tuning technique. The proposed method was evaluated on scans of poinsettias and onions and achieved best average scores of 0.831, 0.839, 0.834, and 0.718 for Precision, Recall, Dice, and IoU, respectively. Further experiments showed reducing the training data to 1 % did not significantly impact the segmentation accuracy. Therefore, the proposed method has promising potential to facilitate RSA analysis with its high utility.
format Article
id doaj-art-2fee4f337a1746bca9b042b7ac157f3f
institution Kabale University
issn 2772-3755
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Smart Agricultural Technology
spelling doaj-art-2fee4f337a1746bca9b042b7ac157f3f2024-12-13T11:08:15ZengElsevierSmart Agricultural Technology2772-37552024-12-019100666Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clusteringMary E. Cassity0Paul C. Bartley1Yin Bao2Department of Biosystems Engineering, Auburn University, Auburn, AL, USADepartment of Horticulture, Auburn University, Auburn, AL, USADepartment of Biosystems Engineering, Auburn University, Auburn, AL, USA; Current address: Department of Plant and Soil Sciences, Department of Mechanical Engineering, University of Delaware, Newark, DE, USA; Corresponding author.X-ray computed tomography (CT) is a powerful tool for in situ plant root system architecture (RSA) characterization. Accurate root segmentation from CT images is integral to studying RSA. Research studies on segmenting roots from CT images have been mainly limited to image processing-based approaches which may require parameter tuning and often lack common segmentation metrics, e.g., Dice and IoU. A recent deep learning approach utilizes a volumetric encoder-decoder network to achieve a high Dice score and IoU. However, training a volumetric model is dependent on the availability of fully annotated scans of the growing medium column, obtaining which can be time-consuming, tedious, and resource intensive. In this study, an efficient method using deep learning-based instance segmentation in conjunction with density-based spatial clustering of applications with noise (DBSCAN)-based filtering was developed and evaluated for two horticultural plant species. A pretrained Mask R-CNN model was fine-tuned on images selected along different axes of the three-dimensional scans to identify the best view selection strategy for volumetric root segmentation. DBSCAN was used to filter noise from the volumetric segmentation with an automated parameter tuning technique. The proposed method was evaluated on scans of poinsettias and onions and achieved best average scores of 0.831, 0.839, 0.834, and 0.718 for Precision, Recall, Dice, and IoU, respectively. Further experiments showed reducing the training data to 1 % did not significantly impact the segmentation accuracy. Therefore, the proposed method has promising potential to facilitate RSA analysis with its high utility.http://www.sciencedirect.com/science/article/pii/S2772375524002715DBSCANInstance segmentationRoot segmentationX-Ray CT
spellingShingle Mary E. Cassity
Paul C. Bartley
Yin Bao
Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
Smart Agricultural Technology
DBSCAN
Instance segmentation
Root segmentation
X-Ray CT
title Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
title_full Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
title_fullStr Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
title_full_unstemmed Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
title_short Root segmentation of horticultural plants in X-Ray CT images by integrating 2D instance segmentation with 3D point cloud clustering
title_sort root segmentation of horticultural plants in x ray ct images by integrating 2d instance segmentation with 3d point cloud clustering
topic DBSCAN
Instance segmentation
Root segmentation
X-Ray CT
url http://www.sciencedirect.com/science/article/pii/S2772375524002715
work_keys_str_mv AT maryecassity rootsegmentationofhorticulturalplantsinxrayctimagesbyintegrating2dinstancesegmentationwith3dpointcloudclustering
AT paulcbartley rootsegmentationofhorticulturalplantsinxrayctimagesbyintegrating2dinstancesegmentationwith3dpointcloudclustering
AT yinbao rootsegmentationofhorticulturalplantsinxrayctimagesbyintegrating2dinstancesegmentationwith3dpointcloudclustering