ForestSense™ — Demystifying the Black Box!

How ForestSense™ sees 80% of your trees and uses them to inventory and spatially illustrate every stand!


Rules of Engagement.
We review with you the specific objectives and guidelines for your project, and we process and import existing data that you have onto our production servers. We will break your property into production blocks and assign to production teams for more efficient processing. Staff teams are trained to recognize and become familiar with the terrain and specifications that characterize your property.

Imagery Collection.
Given your objectives we set exact specifications for the collection of LiDAR and Color Infrared imagery (CIR). We quality check (QC) the LiDAR data for point density, horizontal and vertical accuracy and completeness. We also QC the imagery for radiometric balancing, proper band registration, complete coverage and proper edge matching. Corrections, if required, are completed, and data is uploaded onto our servers for further processing.

Stand Delineation.
The LiDAR and CIR are superimposed so as to provide a highly detailed view of your property so that we can extract subtle changes in forest type, composition, natural boundaries, etc. Often these are much more accurate groupings in terms of key forest attribute than our clients previously were using. We discuss these with you, why we think they are important, and finalize in our system.

Convert Data to Images.
We convert millions of lidar data points into digital, gray scale images that contain tree height information for every visible tree crown on a customer’s property. This provides important information regarding stand density and height distribution, both of which are key inputs to the sample design.

Random Sample Design.
Our biometricians determine the intensity of the sample size, based upon the variations in LiDAR heights, tree density, tree age, imagery characteristics, silvicultural history, site index, and other attributes that may be provided by our clients. Randomly selected stands and fixed area plot locations are determined. Often these plots will include 40 or more trees.

Field Data Collection (Field Forestry).
Our ground-based foresters use the latest GPS and other technologies to correlate LiDAR tree crown locations with actual ground measurements such as DBH, merchantable height, species, grade, etc. ImageTree crews use our proprietary Forest Sense Field Application forestry software to record critical data on a field-grade laptop computer. All plots are marked with both a GPS point and a monument, so that they can be relocated during our subsequent audit process. We ask each crew to review at least 10% of the plots of the other crews to insure the accuracy of data collection.

Creating Tree Polygons—the Containers of Inventory Information.
Using ForestSense software, polygons are created automatically for every visible tree crown in each stand, which usually represents about 80% of all trees. This is an entirely automated process that uses high speed servers within the ImageTree technology infrastructure to create millions of polygons. These polygons are rigorously reviewed by the production quality control team for completeness, and then aligned with CIR imagery, so as to assign each with a species classification.

Biometrics: Extracting the Data.
Once the field data and the tree crown polygons have been completed, the two data sets are spatially merged and prepared for biometric analysis. The relationships among LiDAR, crown polygons, and ground measured data are analyzed using a series of regressions to derive models that predict DBH, species, number of trees, tree grade, etc. for every tree crown polygon.

Database Creation: Compiling the Data.
An inventory database is created using the equations developed during the biometric analysis. The millions of tree crown polygons, now populated with data, are aggregated at the stand level to create stand and stocking tables. We complete our rigorous quality control process before releasing any results to our customers. Our project management team works with you to ensure that our databases are formatted to flow seamlessly into your internal systems.

Delivery of the Results.
The final deliverables for our crown based inventory are:

  1. Raw and stand-based forestry LiDAR data and color infrared imagery.
  2. Geographic information system (GIS) stand polygon layer
  3. Inventory database
  4. Documentation describing inventory methods, summary tables, and statistical accuracy.