Project Overview

In This project, I proposed an algorithm to calculate individual tree (stem) volumes based on LiDAR derived Canopy Height Module (CHM).

The main Workflow steps:

1. Clean-up the CHM dataset: Apply a Gaussian filtering to eliminate spurious values and apply a minimum tree height threshold to filter out intermediate vegetation stages.
2. Delineate individual Tree Crowns using Local Maxima and Watershed segmentation algorithm.
3. Extract explanatory variables (Tree Height & Crown Area) for each tree.
4. Predict DBH using Regression models – from "R.J. Hall et al., 1985: A Comparison of Existing Models for DBH estimation from Large-scale Photos"
5. Estimate Volumes using BC forests 1976 volumes equations.

Dependencies

The following packages need to be installed on your Python Env to run the this code:

• Numpy
• Scipy
• Pandas
• Osgeo (gdal and ogr)
• Scikit-image

Links

Github Repo

Notebook

Slides