Algorithm Details

Supported Sensors
Supported Resolution
5 cm/pixel or less
Other Requirements

- 3 Dimensional outputs needed
- Tree must have leaves at time of survey
- Visible elevation changes in 3D data to delineate tree crowns
Estimated Processing Time
10-12 hours
Georeferenced Image, KML, PDF Map


The purpose of the Tree Crown Delineation algorithm is to automatically outline individual tree canopies and calculate a range of statistics for each tree detected. 

This process is carried out using the 3 Dimensional DSM data detected in the orthomosaic. With the elevation changes seen in the DSM we are able to delineate one tree from the other and create a vector file for each individual tree crown. Attributes are then calculated for each tree detected including- Height, Diameter, and Vegetation Index based statistics. These attributes can be viewed in the Attribute table of the Tree Crown Shapefile produced. 

The algorithm was specifically developed around fruit tree orchards. But we have seen good performance over trees and bushes with observable separation and elevation differences that can be detected in the DSM. It is important to have quality 3 Dimensional outputs for best performance with the Tree Crown Delineation algorithm. The algorithm will not be able to distinguish tree crowns during leaf off season.


The algorithm will process three versions of the output- PDF Map, KML, and Shapefile. 

The most informative output from this is the Shapefile output. This shapefile can be opened in any GIS or Remote Sensing software. Within the attribute table of the Tree Crown Delineation Shapefile you will find a list of statistics calculated against every tree crown detect. This can be used to analyze each tree crown individually through your area of interest. 

  • Diameter- Measurement of tree crown at highest detected diameter
  • Height- Measurement of tree height at peak of tree crown
  • GLI- Mean, Minimum, Maximum, Standard Deviation, Variance of Green Leaf Index algorithm
  • VARI- Mean, Minimum, Maximum, Standard Deviation, Variance of visible atmospheric resistance index algorithm
  • Visual NDVI (NGRDI)- Mean, Minimum, Maximum, Standard Deviation, Variance of Visual NDVI algorithm