Cooks Creek LiDAR DEM

LIDAR Calibration:
ATLIS used Terra Match software (Microstation). In calibration process the laser data position and orientation corrections were calculated using assessing the differences between laser surfaces from overlapping strips and differences between laser surfaces and known ground points.

For misalignment angles between adjacent strips, tie lines were calculated and adjusted. In addition ground control (on flat and solid areas) was incorporated to the solution, in order to solve the vertical bias of the laser data to the bare ground.

Strips misalignment:
Starting Avg. 3D Mismatch [m] 0.018
Final Avg. 3D Mismatch [m] 0.012
Strip misalignment indicates relative accuracy of about 1.5 cm, which is good.

LIDAR Classification:
The Point Cloud was Classified with Microstation v8 (TerraScan software), while using macros that are set-up to measure the angles and distances between points to determine what classification a point should be. The lower points are generally classified as ground returns, with the points above separated in low, medium and high vegetation. Low vegetation is usually between zero and one meter above the ground surface and is not used in the Full Feature product generation. Medium vegetation is typically between one and two meters above ground and high vegetation is everything above two meters. Error points (low point class) are determined to be either high (spikes) or low (pits) outlier points, often beyond 3-sigma from the rest of the data set; clouds, birds, pollution, or noise in the data can cause error points.

After an automated macro is run to determine classes, a manual QC is performed to fine tune and manually correct the classification of points among the different categories. To better understand areas for improvement, the points that are classified as bare earth are extracted and turned into viewable TIN and grid surfaces. These surfaces are inspected for areas that appear rough, artificially flattened or truncated, no data areas, or have other viewable errors.

In cleaning up ground points, the focus is concentrated in areas where few ground points have been left in the bare earth model and the ground appears rough or lower and flatter than it may be in reality. The scarcity of ground points may be a result from no penetration through a dense vegetation layer, water bodies, low reflectivity objects, or too aggressive values with the macro. A manual inspection of these areas plays a major role in resolving any issues or irregularities with the bare earth model.

Classification QC:
Macro level (automatic) classification was conducted for Noise (Low and High), Ground (Class 2), Water (Class 6) and Vegetations (Class ). After automatic classifications, the whole dataset was checked in accordance to DEM to identify misclassification issues. Objects such bridges and low/high noise were identified and corrected. Profile Checks: (Ground - Vegetation separation). DEM: DEM was interpolated from the bare ground points of the classified LAS data at a 1m spacing and exported as Arc ASCII Raster and ASCII text file.

Process Date Feb 24, 25, 26, 27, 29, 2016