How to Create 3D Terrain Models Using LiDAR Scans

Hey everyone, here is a 3D terrain scan I captured using the SH 20 lidar scanner. This 2,000 square meter terrain scan holds more than 9 million points and we captured it in under 10 minutes. That's what makes SLAM lidar such a powerful tool for building high precision topographic models. Today, I'll be showing you how to turn a lidar scan into a thin surface, a contour plan, and finally, a detailed 3D terrain model. Among topographic maps, the 3D terrain model stands out as the most useful. In fact, when I take on lidar scan projects, architects almost always prefer the 3D terrain model over a simple contour plan. This terrain allows designers to model directly over it, and thanks to the CAD format, the 2, 000 square meter file is only 5 MB in size. This is the original scan data. The vegetation here is so dense that it blocked the GPS signal, making RTK unusable. On projects like this, customers typically request georeferenced data. Fortunately, with the SH 20, you can integrate GPS coordinates into the scan as a workaround. Step one, place four checkerboards in a clear open space close to the site. The next step is to scan the dense vegetation area in regular loops and then move into the open space to capture the ground control points. Capturing each control point takes about 10 seconds, which means you will lose around 1 minute of scanning time in total. Then loop back to your starting point to close the trajectory. When you complete multiple loop closures like in this case with two loops, the SLAM accuracy actually surpass scanning with the RTK. Next, in the open area where you set up the checkerboard, perform a second scan with RTK, making sure all the boards are fully captured. Next, remap the scan using RTK first, then open the measurement tab. I don't have the real sample to show here. This is just demonstration of how to copy GPS coordinate in CH Studio. Simply click on the center of the checkerboard with your mouse, copy the coordinate, and paste it into Notepad. Next, remap the high vegetation area scan using control point calibration. Paste the copied GPS coordinates into the input box and run the calculation. Once that's done, the terrain scan will include GPS data. This method gives you the highest accuracy from SLAM loop closures while still providing GPS coordinates for georeferencing. Before creating the thin surface, you will need to trim away all the vegetation. I usually do this using cropping tool in CloudCompare.