The Stephen F. Austin Experimental Forest in east Texas is our research area for this project. The first objective of our research is to conduct accuracy assessment by comparing ground measured data to lidar derived data. Different combinations of data processing algorithms and data sources for lidar derivation will be used in order to see if one is significantly more accurate than another. Eventually, a fine tuned model will be built as the best estimation of the forest through lidar remote sensing.
The second objective is to rastering lidar point cloud data for the fusion of multispectral imagery with the attempt to increasing cover type classification accuracy. Large-area forest inventory relies on accurately delineating forest cover types through remote sensing. Lidar derived raster data are expected to provide additional perspectives for distinguishing forest features due to its 3D structure.
Ultimately, we will integrate all of the data into a geodatabase where the position of each individual tree (x, y coordinate) and its associated attributes (species, DBH, height, basal area, crown diameter, etc.) are stored. Through modeling, different forest practices such as thinning can be simulated before a decision is made.
Profile view of Lidar point cloud
Lidar point cloud of the SFA Experimental Forest
Reseaching at the SFA Experimental Forest