Reality Capture for Outdoor Terrain Analysis
|The achilles heel of terrain analysis has historically been access to maps however with the Stencil and Contour it is now possible to perform reality captures within minutes.
The rest of this article documents the process we followed to perform a basic terrain ruggedness analysis on a suburban park.
Collection of the data is now made extremely simple, please contact us if you would like a demo. The reality capture shown below took approximately 15 minutes.
|Before heading out to perform the capture, we decided on the optimal path to be taken. This ensures that full coverage of the landscape is achieved maximising the poses that feature rich landmarks are captured from. Once we successfully captured the area, we headed back to the office to check out the result.
The first step was to review the data in CloudCompare. CloudCompare is a free and open source utility for viewing and editing point clouds. It is a significantly capable piece of software given its price tag. CloudCompare also comes installed in the Contour by default so you can review point clouds in the field before heading off.
|Once the point cloud had been verified in CloudCompare we imported it into SAGA. SAGA is a free and open source geospatial processing software aimed at the automation of point cloud processing. It is an incredibly powerful tool that encompasses a lot of functionality in the geospatial domain. We decided to use it for this article.
After the point cloud had been successfully loaded into SAGA we went ahead and created a digitial terrain model (DTM) shown below.
|The next step, performing the terrain ruggedness analysis, is straight forwards. SAGA provides a number of methodologies for terrain modelling and we picked one of the more basic algorithms to prove the concept. Depending on the real-world application we can help our customers to finely tune their workflows to suit their needs.|
|Interpretation of the data matches the results closely. Referring to the legend, areas of denser tree foliage have a more rugged terrain and open area’s which are well kept show up clean. Applying change-set detection over the top of this would yield some interesting result with respect to the management of foliage for public infrastructure. Cost functions can also be applied to the data so that the process of capture to analysis can be fully automated depending on the application at hand.|