![]() |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
In addition to its convenient user interface, SLIM3D offers the possibility to run the processing steps almost entirely automatically via Visual Basic® Script. Interactive control is only required for adjusting the parameters if a new class of objects or a new 3D sensor is used. Moreover, SLIM3D is fully controllable by DCOM function calls and thus seamlessly integrable into other programs. System requirements: SLIM3D runs under Windows® NT/2000/XP and requires at least a Pentium® 4 processor with 1024 MB RAM and a minimum of 300 MB free hard disk space. Further details can be found in the PDF file of the product flyer.
|
SLIM3D is a program package for processing the data from optical 3D measuring systems for |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| In registration, the depth images of different views
of an object are aligned to one another without needing to know the
scanning position or the sensor type. An automatic coarse registration
step makes it possible to place the sensor in any desired position
around the object and so ensure that there is no missing data caused
by occlusion or shadowing. If automatic registration does not yield
satisfying results, an interactive alternative can be used. For this,
corresponding features such as object corners or significant textures
are marked by the user in the overlapping areas of different scans. Remaining surface deviations in the overlap regions are then minimized by a fine registration procedure. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
The aligned depth images are then converted, fully automatically, into triangle meshes and cleaned of noise, aliasing and outliers. Next, the meshes for the different views are merged into a single surface description. Any interference that appears owing to registration or calibration errors is then filtered out. Finally, any gaps that are present are closed using interpolation in a fully automated process, and curvature-dependent mesh thinning is carried out. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||