Limits...
Automated interpretation of 3D laserscanned point clouds for plant organ segmentation.

Wahabzada M, Paulus S, Kersting K, Mahlein AK - BMC Bioinformatics (2015)

Bottom Line: We investigate the performance of the resulting approaches in the practical context of grouping 3D point clouds and demonstrate empirically that they lead to clustering results with high accuracy for monocotyledonous and dicotyledonous plant species with diverse shoot architecture.This approach is applicable to different plant species with high accuracy.The analysis cascade can be implemented in future high-throughput phenotyping scenarios and will support the evaluation of the performance of different plant genotypes exposed to stress or in different environmental scenarios.

View Article: PubMed Central - PubMed

Affiliation: INRES-Phytomedicine, University of Bonn, Meckenheimer Allee 166a, Bonn, 53115, Germany. mirwaes@uni-bonn.de.

ABSTRACT

Background: Plant organ segmentation from 3D point clouds is a relevant task for plant phenotyping and plant growth observation. Automated solutions are required to increase the efficiency of recent high-throughput plant phenotyping pipelines. However, plant geometrical properties vary with time, among observation scales and different plant types. The main objective of the present research is to develop a fully automated, fast and reliable data driven approach for plant organ segmentation.

Results: The automated segmentation of plant organs using unsupervised, clustering methods is crucial in cases where the goal is to get fast insights into the data or no labeled data is available or costly to achieve. For this we propose and compare data driven approaches that are easy-to-realize and make the use of standard algorithms possible. Since normalized histograms, acquired from 3D point clouds, can be seen as samples from a probability simplex, we propose to map the data from the simplex space into Euclidean space using Aitchisons log ratio transformation, or into the positive quadrant of the unit sphere using square root transformation. This, in turn, paves the way to a wide range of commonly used analysis techniques that are based on measuring the similarities between data points using Euclidean distance. We investigate the performance of the resulting approaches in the practical context of grouping 3D point clouds and demonstrate empirically that they lead to clustering results with high accuracy for monocotyledonous and dicotyledonous plant species with diverse shoot architecture.

Conclusion: An automated segmentation of 3D point clouds is demonstrated in the present work. Within seconds first insights into plant data can be deviated - even from non-labelled data. This approach is applicable to different plant species with high accuracy. The analysis cascade can be implemented in future high-throughput phenotyping scenarios and will support the evaluation of the performance of different plant genotypes exposed to stress or in different environmental scenarios.

No MeSH data available.


Related in: MedlinePlus

Quantative results showing the F-measure and entropy values as a function of number of clusters. The F-measure results (top row) show a better performance for Algorithm 1 using data mappings (HC) than those for k-means (KM) clustering on normalized histograms directly. This is also captured by the entropy values (middle row), as it considers the distributions of different labels within the clusters. The lower value, the more the clusters are dominated by histograms of a particular label, and therefore the better the clustering. For all methods the algorithm required only few minutes per run and the number of cluster (bottom row)
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4528849&req=5

Fig2: Quantative results showing the F-measure and entropy values as a function of number of clusters. The F-measure results (top row) show a better performance for Algorithm 1 using data mappings (HC) than those for k-means (KM) clustering on normalized histograms directly. This is also captured by the entropy values (middle row), as it considers the distributions of different labels within the clusters. The lower value, the more the clusters are dominated by histograms of a particular label, and therefore the better the clustering. For all methods the algorithm required only few minutes per run and the number of cluster (bottom row)

Mentions: Figure 2 summarizes the results for the grapevine, wheat and barley datasets where manual annotations were given. For the grapevine dataset consisting of stem and leaves, as well as the wheat dataset with leaves, stem and ears, we computed separately the clusters, whereas for clustering the barley data containing of leaves and stems, we considered the histograms of all three datasets together. Since k-means is based on random initialization, which consequently can lead to different clusterings, each experiments was repeated five times to provide reliable results. We report averaged values of all runs as a function of number of clusters.Fig. 2


Automated interpretation of 3D laserscanned point clouds for plant organ segmentation.

Wahabzada M, Paulus S, Kersting K, Mahlein AK - BMC Bioinformatics (2015)

Quantative results showing the F-measure and entropy values as a function of number of clusters. The F-measure results (top row) show a better performance for Algorithm 1 using data mappings (HC) than those for k-means (KM) clustering on normalized histograms directly. This is also captured by the entropy values (middle row), as it considers the distributions of different labels within the clusters. The lower value, the more the clusters are dominated by histograms of a particular label, and therefore the better the clustering. For all methods the algorithm required only few minutes per run and the number of cluster (bottom row)
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4528849&req=5

Fig2: Quantative results showing the F-measure and entropy values as a function of number of clusters. The F-measure results (top row) show a better performance for Algorithm 1 using data mappings (HC) than those for k-means (KM) clustering on normalized histograms directly. This is also captured by the entropy values (middle row), as it considers the distributions of different labels within the clusters. The lower value, the more the clusters are dominated by histograms of a particular label, and therefore the better the clustering. For all methods the algorithm required only few minutes per run and the number of cluster (bottom row)
Mentions: Figure 2 summarizes the results for the grapevine, wheat and barley datasets where manual annotations were given. For the grapevine dataset consisting of stem and leaves, as well as the wheat dataset with leaves, stem and ears, we computed separately the clusters, whereas for clustering the barley data containing of leaves and stems, we considered the histograms of all three datasets together. Since k-means is based on random initialization, which consequently can lead to different clusterings, each experiments was repeated five times to provide reliable results. We report averaged values of all runs as a function of number of clusters.Fig. 2

Bottom Line: We investigate the performance of the resulting approaches in the practical context of grouping 3D point clouds and demonstrate empirically that they lead to clustering results with high accuracy for monocotyledonous and dicotyledonous plant species with diverse shoot architecture.This approach is applicable to different plant species with high accuracy.The analysis cascade can be implemented in future high-throughput phenotyping scenarios and will support the evaluation of the performance of different plant genotypes exposed to stress or in different environmental scenarios.

View Article: PubMed Central - PubMed

Affiliation: INRES-Phytomedicine, University of Bonn, Meckenheimer Allee 166a, Bonn, 53115, Germany. mirwaes@uni-bonn.de.

ABSTRACT

Background: Plant organ segmentation from 3D point clouds is a relevant task for plant phenotyping and plant growth observation. Automated solutions are required to increase the efficiency of recent high-throughput plant phenotyping pipelines. However, plant geometrical properties vary with time, among observation scales and different plant types. The main objective of the present research is to develop a fully automated, fast and reliable data driven approach for plant organ segmentation.

Results: The automated segmentation of plant organs using unsupervised, clustering methods is crucial in cases where the goal is to get fast insights into the data or no labeled data is available or costly to achieve. For this we propose and compare data driven approaches that are easy-to-realize and make the use of standard algorithms possible. Since normalized histograms, acquired from 3D point clouds, can be seen as samples from a probability simplex, we propose to map the data from the simplex space into Euclidean space using Aitchisons log ratio transformation, or into the positive quadrant of the unit sphere using square root transformation. This, in turn, paves the way to a wide range of commonly used analysis techniques that are based on measuring the similarities between data points using Euclidean distance. We investigate the performance of the resulting approaches in the practical context of grouping 3D point clouds and demonstrate empirically that they lead to clustering results with high accuracy for monocotyledonous and dicotyledonous plant species with diverse shoot architecture.

Conclusion: An automated segmentation of 3D point clouds is demonstrated in the present work. Within seconds first insights into plant data can be deviated - even from non-labelled data. This approach is applicable to different plant species with high accuracy. The analysis cascade can be implemented in future high-throughput phenotyping scenarios and will support the evaluation of the performance of different plant genotypes exposed to stress or in different environmental scenarios.

No MeSH data available.


Related in: MedlinePlus