Surfaces from Points | Spatial Cognition

Reconstructing Surfaces

In order to build traditional 3D models, perform analysis tasks, or visualize high-resolution imagery on sparse point clouds, we need to reconstruct surfaces. We have a novel, real-time process to create detailed surfaces from very large point datasets.

This is not a preprocessing step. Our products automatically build surfaces with multiple levels-of-detail from any available point data, including combining data from different point clouds and incorporating streaming data as they come in from live sensors. Because surfaces are generated on-demand for the areas needed, this process works with terabytes of point data that can span hundreds of square miles but can still generate very fine detailed geometry.

Unlike many other products, our solution works for both interiors and exteriors. It identifies small details in 3D objects and extracts them from the “background scene”. It also provides real-time, high-accuracy analysis of the point data for functions like measurement tools.