Recap

In the previous tutorial section, we detected tree stems, segmented individual trees, and reconstructed their branching structures using the Tree Shape Metrics 3D tool. We applied this workflow to an example dataset containing six trees and derived structural metrics such as wood volume and biomass.

Before extending this workflow to a larger area of interest, we will explore two additional applications of the Tree Shape Metrics 3D tool:

1. Stem reconstruction and straightness: We will examine how to improve stem reconstruction when our main interest is stem geometry and straightness, particularly for assessing sawlog quality.
2. Irregular stem cross-sections: We will explore how to reconstruct old, solitary trees, such as those found in urban parks, which often have irregular or noncircular stem cross-sections.

These examples will help us understand how to adapt the reconstruction workflow to different tree structures and analysis objectives.

Open a New Project

To get started, we first need to create a new, empty LIS Pro 3D project.

From the top menu, select File > Project > New Project.

Download the Prepared Trees

Download the prepared tree datasets from here.

Download the prepared trees folder as a ZIP archive and extract it into your project folder.

The extracted folder contains two example trees and their corresponding tree position files:

Load the Coniferous Tree into the GUI

Load the picea_abies.sg-pts-z point cloud and the picea_abies_tree_positions.shp shapefile into your empty project.

Run Tree Shape Metrics 3D – Basic Mode

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

After running the tool, four new shapes layers appear in the Data tab:

• picea_abies_metrics: A point shapes layer containing aggregated tree-level metrics, such as tree dimensions, crown characteristics, wood volume, and diameter at breast height (DBH).
• picea_abies_skel: A line shapes layer representing the detailed tree skeleton as a graph. Each graph edge represents a section of the tree and contains geometric, topological, and pipe model attributes.
• picea_abies_axes: A simplified representation of the tree skeleton in which graph edges belonging to the same branch are merged into a single feature. Attributes are reported per branch rather than per graph edge.
• picea_abies_pipe: A polygon shapes layer representing the reconstructed pipe model. It contains the same attribute set as the detailed skeleton.

Additionally, a new point cloud dataset is created with additional attributes compared to the original input point cloud.

Inspect the Results in 3D

Open the input point cloud and the reconstructed pipe model (picea_abies_pipe) in the Point Cloud Editor to inspect the results in 3D.

Tool: Point Cloud Editor
Geoprocessing → LIS Pro 3D → Point Cloud Editor   //    Tools → LIS Pro 3D → Point Cloud Editor

Press P to switch to parallel projection.

Press 2 to colorize the point cloud by Intensity.

You may notice that the reconstructed trunk geometry is slightly offset from the point cloud and appears somewhat irregular or wavy.

Run Tree Shape Metrics 3D – With Fine Adjustment

To improve the geometric accuracy of the reconstructed trunk, we run the Tree Shape Metrics 3D tool again, this time enabling fine adjustment.

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

After running the tool, you can see that the reconstructed model is now better aligned with the point cloud.

However, some misalignment may still be visible in the upper crown. This is typical of TLS datasets, where point density often decreases and data gaps become more frequent toward the upper canopy.

To improve reconstruction stability, particularly for monopodial trees with a straight, dominant main stem, we recommend a two-step approach: reconstruct the trunk first, followed by the remaining branching structure.

Run Tree Trunk Slicing

Now, we run the Tree Trunk Slicing tool to reconstruct the main stem geometry before refining the remaining branching structure.

Tool: Tree Trunk Slicing
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

The tool analyzes 96 horizontal slices and attempts to estimate the stem diameter and center position at each slice height.

After running the tool, three output datasets are generated:

• picea_abies_axes: A line shapes layer representing the reconstructed medial axis of the trunk. Note that this dataset has the same name as the branch axes layer generated earlier by Tree Shape Metrics 3D.
• picea_abies_crosssections: A point shapes layer containing the center positions of the detected stem cross-sections. It also stores quality metrics for each fitted circle, such as Circle Completeness and Goodness of Fit, at the corresponding slice height.
• picea_abies_stem_pipes: A polygon shapes layer representing the reconstructed stem-only pipe model, which can be inspected alongside the point cloud.

View the Stem Pipes in 3D

Tool: Point Cloud Editor
Geoprocessing → LIS Pro 3D → Point Cloud Editor   //    Tools → LIS Pro 3D → Point Cloud Editor

The picea_abies_stem_pipes dataset visualizes the reconstructed trunk axis and the fitted diameters at each slice height as a polygonal mesh.

If the reconstructed trunk axis is sufficiently accurate, we can use the picea_abies_axes dataset, which represents the medial axis as a line shapes layer, to guide the reconstruction of the complete tree structure.

To do this, we run the Tree Shape Metrics 3D tool again and provide the previously generated picea_abies_axes dataset from Tree Trunk Slicing as the trunk axis input.

This allows the algorithm to prioritize the reconstructed main stem when deriving the remaining branching structure.

Run Tree Shape Metrics 3D – With Trunk Axis

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

The existing reconstruction results are updated after running the tool.

Let’s inspect the updated reconstruction in the Point Cloud Editor again.

Tool: Point Cloud Editor
Geoprocessing → LIS Pro 3D → Point Cloud Editor   //    Tools → LIS Pro 3D → Point Cloud Editor

The reconstructed pipe model now follows the observed stem geometry more closely, particularly in areas where the previous reconstruction deviated from the point cloud.

Log Classes

Another output of the Tree Shape Metrics 3D tool is the classification of stem sections into logs based on their length and straightness.

In the Point Cloud Editor, open Advanced Settings. In the 3D Shapes Layer section set the Color Field to SweepClass and Mode Color Range to Minimum/Maximum.

Click Okay to close the dialog.

All branches in the reconstructed model are colorized dark blue, indicating a SweepClass value of 0 (no usable log).

The main stem is divided into two log sections, excluding the lowest and uppermost parts of the tree.

Logs are classified by length (long or short) and straightness, resulting in four classes:

1. Class 1: Long logs with the best straightness.
2. Class 2: Long logs with moderate straightness.
3. Class 3: Short logs with good straightness.
4. Class 4: Short logs with moderate straightness.

To configure the log classification parameters, expand the Log Classification settings at the bottom of the Tree Shape Metrics 3D parameter interface.

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

The goal is to identify logs with target lengths of 6 m (long logs) and 4 m (short logs), with long logs preferred over short logs.

Rerun Tree Shape Metrics 3D with Different Log Classes

Let’s run Tree Shape Metrics 3D again, this time slightly adjusting the two target log lengths in the Log Classification settings.

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

The updated classification identifies one 8 m log and two additional logs, each longer than 2 m.

Get Tree and Stem Info

Let’s create tabular data from the information we have derived for the tree.

Tool: Extract Tree Information 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

As a result, we obtain three tables:

1. picea_abies_metrics_info
2. picea_abies_metrics_stem
3. picea_abies_metrics_branches

Reconstructing a Solitary Park Tree

So far, we have focused on reconstructing the geometry of a conifer with a relatively straight, dominant main stem. However, older deciduous trees often have much more complex structures, including broad crowns, large spreading branches, and irregular trunk shapes.

In this final example, we will explore how Tree Shape Metrics 3D handles these challenges using a solitary Norway maple (Acer platanoides) growing in a park. We will pay particular attention to the reconstruction of noncircular trunk cross-sections and the branching structure.

Open an Empty Project

To conclude our detailed exploration of the Tree Shape Metrics 3D tool, we will start with a new, empty project.

Select File > Project > New Project.

Load Deciduous Tree into GUI

In the prepared trees folder, you will find the following datasets for the Norway maple:

Run Tree Shape Metrics 3D Tool (with Fine Adjustment)

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

The park tree has an impressive architecture and a large trunk diameter at its base. Its trunk geometry is particularly complex, with highly irregular, noncircular cross-sections near the base and around the main fork.

Representing such a large tree using only circular cross-sections can oversimplify its geometry and may lead to inaccurate reconstruction results.

As you can see, the reconstructed branch axes do not align particularly well with the point cloud, especially around the main fork.

To improve the fit, we can use an approach based on amorphous cross-sections, allowing the reconstructed model to better represent the irregular geometry observed in the point cloud.

Run Tree Shape Metrics 3D with Amorphous Trunk Base

Tool: Tree Shape Metrics 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

We can see that the reconstructed model now aligns much more closely with the point cloud, particularly around the trunk base and main fork.

Viewing the Basic Results

Besides the pipe model (acer_platanoides_pipe), which is primarily intended for visualization, the detailed geometric and structural information is stored in three additional datasets:

• acer_platanoides_rings
• acer_platanoides_axes
• acer_platanoides_skel

Generate Basic Tree Information

In addition to these detailed reconstruction outputs, we can extract key tree metrics and structural information from the reconstructed datasets and organize them into summary tables.

Tool: Extract Tree Information 3D
Geoprocessing → LIS Pro 3D → Forestry → Point Cloud   //    Tools → LIS Pro 3D → Forestry

As a result, we obtain two tables:

1. acer_platanoides_metrics_info
2. acer_platanoides_metrics_stem

Summary

In this tutorial section, we explored advanced applications of the Tree Shape Metrics 3D tool using two trees with very different growth forms and reconstruction challenges.

1. Coniferous trees with a dominant, relatively straight stem: Using a Norway spruce (Picea abies), we examined how stem reconstruction can be improved by combining Tree Trunk Slicing with Tree Shape Metrics 3D. We also explored stem straightness analysis, log classification, and the identification of potentially usable sawlogs based on their length and allowable sweep.

2. Old, solitary broadleaf trees with complex stem geometry: Using a Norway maple (Acer platanoides), we demonstrated how amorphous cross-sections can improve the reconstruction of irregular trunk bases and major forks. We also examined the resulting rings, branch axes, and tree skeleton, including the hierarchical representation of branching architecture and the derivation of wood volume and aboveground woody biomass estimates.

Together, these examples demonstrate how Tree Shape Metrics 3D can be adapted to different tree architectures and analytical objectives, ranging from stem quality assessment and potential sawlog identification to detailed structural analysis and biomass estimation.

< Previous section    Next section >