Page 26 - 3DF Zephyr

process tougher to noise, whereas a higher value will keep the detail although the point cloud will be more exposed to reconstruction noises (if there are low-quality photos for instance). Notice that using high value settings will affect your hardware memory.  Voting type: calculate a hierarchy for the dense point cloud when noise reduction is applied. It is recommended to select the hierarchical option.  Shape from Silhouette: (Zephyr Pro only) it is an image-based modeling technique that uses the profile, or silhouette, of the subject to create a virtual hull containing the subject. You must create the masks for all the pictures before running that. This technique can be quite worthwhile when you need to deal with transparent or reflective surfaces which can not be removed with 3DF Masquerade. Once the parameters are chosen you can start the process by clicking on Run button. The dense point cloud generation requires the most of hardware capabilities along with the cameras orientation phase. Don't forget that you can reconstruct and add more dense point clouds in the same Zephyr project even exploiting different parameters or reconstruction volumes. A further description of the parameters above is available here: http://www.3dflow.net/3df-zephyr-parameters-tuning-guide/ Note: if dense point clouds are your goal, you may also want to consider generating a mesh. 3DF Zephyr features the photoconsistent mesh optimization, which allows to improve the geometry based on the photoconsistency of each point that has been generated. Since version 3.100, you can generate a mesh and then turn that mesh back into a dense point cloud. Extracting a mesh Click on Workflow > Mesh Extraction (13) to start the mesh generation phase.