Calculating Tree Inventory from Aerial Lidar

Lidar and other remote sensing methods have become a staple method for measuring modern forests. To penetrate dense canopies, foresters use lidar data in Global Mapper Pro to map bare ground elevation values and visualize forest density, crown coverage, and tree heights. Timber cruises for manual measurements take manpower and time. Calculating a tree inventory from aerial lidar, as well as other measures of life such as midstory cover and exposed ground, helps foresters have a better understanding of existing stand dynamics at a fraction of the labor costs.

This tree inventory workflow focuses on calculating stand count, individual tree height, and location. You’ll benefit from using lidar data that has a high enough density that separate crowns can be discerned, as well as the ground beneath. Need lidar data? Check out this blog: Free Elevation Data Sources to Use in Global Mapper. 

There are two main steps, both executed from the Automatic Point Cloud Analysis tool: 

1. Classify the point cloud to identify tree points. 
2. Extract the point cloud into individual tree features. 

Lidar Classification: Identifying Noise, Ground, and Vegetation Points

Point classification separates vegetation from ground points, an essential step in the processing of forest data. First, classify the Noise and Ground points within the cloud:

• Removing any outliers with the Noise Classification tool will create a more accurate classification overall. 
• An accurate Ground Classification will provide solid groundwork for measuring tree heights during vegetation classification. 

Next, classify the vegetation points. There are two options in the automatic classification tool, Grid, and the more commonly used Max Likelihood.

Tree Classification: Vegetation 

How does Vegetation Classification work? Global Mapper’s Max likelihood classification tool uses Segmentation to identify individual trees in the lidar. Segmentation, to put it simply, is a method of creating clusters of points that represent an individual feature. Each cluster of points, so each detected tree, is visualized using a random color, as shown in the image below. 

Vegetation Classification Settings

While the default settings can identify trees in most situations, for best results, adjust the tool’s settings to tailor the tool to fit your forest type and data. As always, don’t forget to consider the Resolution. 

Vegetation Settings: For Max Likelihood, there is only one setting specific to Vegetation point classification. 

• Classify Individual Trees by Height: Enable this option to assign trees a lidar classification based on height (low, medium, or high vegetation) as determined by the Shared Settings. Leaving this option unchecked will assign all detected vegetation points #4 medium classification. 

Shared Settings: Because these settings reflect the forest structure and not the lidar processing, these shared settings will be used during Classification and Extraction. 

• Tree Spread: These settings refer to the minimum and maximum width (edge-to-edge) of a single tree canopy. This helps in differentiating clumps of trees by informing Global Mapper in what to expect the smallest and largest tree canopies to be in this forest.  
• Min Height Above Ground: Vegetation points below this value will be ignored. This helps to distinguish the understory from the trees.
• Medium Vegetation Threshold: Trees with heights detected between this value and the Max Vegetation value will be classified as Medium Classification. Trees below this value will be classified as Low Vegetation. 
• Max Vegetation Threshold: Trees taller than this value will be classified as High Vegetation. 

Vegetation Extraction 

The Feature Extraction tool creates vector features from classified points. This is where the tree inventory can be created for easy view and export. For trees, the Vegetation Extraction tool generates:

• A vector point feature located at the top of each detected tree 
• An area feature to outline each tree crown  
• A 3D mesh layer generated from the segmented points

A tree point is created at the center of each segment and contains the tree’s measured attributes, such as height, canopy spread, and classification. To see the count of features, look at the Extracted Trees layer in the control center, or open the attribute table to see the feature count in the bottom right-hand corner. This point layer can be exported to many different file types, including a CSV, and as a GMMP compatible with Global Mapper Mobile for viewing and editing in the field. 

Try this for yourself in Global Mapper Pro with a free 14-day trial! If you have any questions, please contact us.

Other blogs on Forestry workflows in Global Mapper Pro:
Measuring Tree Height from Lidar

Create Canopy Height Models from Lidar in Global Mapper Pro

Identifying Change in Tree Height by Comparing Point Clouds to Terrain