Point cloud

Point cloud, point clusters or cluster is called a point set ( a set of discrete values), the full - is located in a defined environment - or predominantly.

Use

Computer Graphics

In 3D computer graphics and computational geometry a point cloud means a list of 3D Cartesian coordinates. These can be created by a 3D modeling tool and means of scanning of objects or surfaces by systems such as coordinate measuring machines or groping 3D scanner. Optical scanners are divided into laser technology, which operate on the principle of triangulation, and normal light scanners ( " coded -light" ) work on the strip -light method. A summary overview of the diversity and performance of current optical scanning methods and the further processing of the resulting 3D-Daten/Punktwolke are, for example, C. Teutsch.

A point cloud is used:

  • For the visualization to represent scanned objects and surfaces on the computer;
  • For modeling; while the original (mostly) closed surface is restored by means of surface reconstruction. The resulting surfaces are usually made of polygons (see also Meshing );
  • As the basis for geometric calculations, eg for measuring human ( body scanning ) or objects;
  • To identify people or objects.

CAD

Body

In the area of ​​CAD point clouds are often used to read in scanned design objects in the CAD system. For demanding forms (eg, automobile body ) a " clay model " ( " clay model " ) is often created. With draw knife and other hand tools, the molds are created from a modeling and scanned it. In symmetrical components (eg bonnet ) just one side (eg only the left side ) is modeled. This is then scanned with tactile, measured or optical scanners. The resulting 3D geometry is first only from points in space ( Y, X, Z coordinates). This point cloud is either read into a special software for " reverse engineering " or read in some cases, directly into the CAD software. The often customary conversion of points to simple surface meshes, as it is usually sufficient in computer games is not enough in the automotive industry. Here Bezier and NURBS surfaces are defined by the points, while balanced with sophisticated filtering methods possible measurement error. The resulting surfaces have high quality thanks to the reworking of an experienced CAD technician exact area boundaries and well-defined transitions (tangential and mostly continuous curve ) to the adjacent surfaces. Which emanated from the " clay model " CAD model can now be relatively easily mirrored in CAD. You now have an exactly described model of a car body (or parts thereof). For this 3D model now another " clay model " of modeling clay in 1:1 scale for further assessment can be milled. Minor design changes can be made to the model again, be re- scanned and re- inserted again with reverse engineering in the now new, reconditioned CAD model. These steps can be repeated frequently and also explain the high costs and time involved in developing a new body ( or interior ). There are efforts to increasingly reflect these steps in specific visualization programs, but you do not want to give up the clay model.

Data formats

To import point clouds in the CAD software mainly following file formats are used: IGES, STEP, DXF, IV, STL, STL -A ( ASCII), STL -B (binary ), WRL, RCP and RCS.

Statistics

In statistics and exploratory data analysis point clouds for the graphical representation of bivariate correlations (cf. scatter plot, correlation). They allow you to gain easy visual impression of the direction and tightness of the connection and to identify outliers in the data set.

665213
de