Procedure

1. With a 3D scanner the object is scanned from several directions.

   Scanning of the object with the 3D sensor.

   Resulting Scan in the QTSculptor software.

   
   

2. The object can be easily segmented from the background.

   Scan with undesired background information.

   The background information can easily be masked out by marking the relevant regions with the mouse.

   
   

3. Using partly overlapping views the 3D registration determines the exact relative orientation of the different views in subpixel accuracy.

   To register two scans three tie points have to be set with the mouse (green crosses). The system automatically determines the exact tie points (purple crosses) and brings the scans into a common coordinate system. The blue cross shows the corresponding point to the one unter the mouse pointer in the other scan.

   
   

4. The preview of the 3D model allows an easy interactive view planning.

   Integrated preview of the point cloud. Regions where no data is available are clearly visible.

   
   

5. Scans belonging to different views are integrated into a polygonal model of the object.

   QTSculptor calculates a triangle mesh of the object in a defined resolution using the point cloud. This mesh can be displayed in an integrated 3D viewer.

   
   

6. The size of the triangles is adapted to the complexity of the surface.

   Model before subdivision; all triangles have approximately the same size.

   In all regions where the complexity is high and the triangles are too big to give a good representation quality of the object QTSculptor subdivides the triangles to elaborate more detail.

   
   

7. Missing parts are interpolated in the model.

   Model with missing parts.

   The missing parts have been interpolated and the model is now waterproof.

   
   

8. Interesting parts in the mesh can be interactively selected and precisely edited.

   Parts of the object can be selected (red region) and edited selectively afterwards with elaborate algorithms.

   
   

9. If desired the acquired geometry can be textured from photographs of the object.

   Color photos taken with a conventional digital camera can be registered with the mesh to texture the model.