Our approach takes images from smart cameras, and generates 3D reconstruction models using structure from motion algorithms. Since end of 2017, our models are generated using COLMAP.


For the data collection phase, the only necessary equipment is the digital camera. Additionally, we recommend placing some markers in the scane, for the purpose of later on scaling and measurement.

For image processing, as shown in the diagram above, multiple toolkits are available to reconstruct Poind Clouds, and the solid models. Pick any toolkit combinations that work best for you.

To render and analyze the reconstruction result, you only need some free third party applications like VisualSFM, Meshlab, and CloudCompare. They are lightweight and easy to install. Read on click here.

The same free applications can be used for creating the reconstruction models. In the case of using CMPMVS application for mesh generation, a NVidia CUDA-enabled GPU may be required.