This news article was originally written in Spanish. It has been automatically translated for your convenience. Reasonable efforts have been made to provide an accurate translation, however, no automated translation is perfect nor is it intended to replace a human translator. The original article in Spanish can be viewed at Digitalización 3D

3D digitising

Miguel Ángel León (AIJU)01/09/2003

The term "3D digitising", appeared at the end of the 1990s, designates three dimensional measures acquisition techniques. 3D digitizers allow us to build CAD models from physical objects.

The needs that tries to answer the 3D scanning and 3D are ancient: control of industrial quality, verification of industrial parts, design of molds, etc. New 3D digitalization techniques allow us to capture the points of measure with more precision and speed, and can be used on different types of objects, of different dimensions, geometry and textures.

There are various 3D scanning systems, which can mainly be divided into two groups: contact or systems without contact with the object to scan.

Scanning by contact

These 3D digitising systems are the oldest. They are mainly used for dimensional verification of industrial parts (quality control). The coordinates of the points due to the displacement of a point on the surface are obtained with these systems to digitize.

Currently, there are heads of digitalization in continuous (SP 600 of Renishaw, see photo). Data acquisition speed increases significantly with respect to conventional headsets, because they are not separated from the surface to digitized.

Another kind of 3D scanning contact system are articulated arms of manual operation. These have a high precision; but on the other hand they have a very low data acquisition speed, because you need to manually bring the tip to each point you want to scan, with which the digitization of a medium-sized object is very expensive.

To use these systems by contact, requires that the pieces have enough stiffness so they deformed does not contact the tip and because of the geometry of the tips, it is impossible to digitize some slots and Interior angles; but they have a high resolution.

Scanning non-contact

The main advantage of the non-contact 3D digitizers is to have a speed of acquisition of data very superior to the digitizers with contact. We can divide digitization techniques without contact in two main groups:

Techniques of passive vision

The human visual system allows to obtain information of depth through the fusion of two monoculares scenes, which are the scenes that capture each of our eyes. This system allows us to "see in 3D" and is known as a principle of stereoscopic vision. This principle is based on two points of view of an object to use to find the three-dimensional coordinates. To determine the position of a point from two images it is necessary to have a model of the optical system used.

This general principle may improve with more elaborate camera models or using two cameras, what is known with the name of Photogrammetry. The stereoscopic vision presents the interest of being able to give the position in 3D of points on surfaces regardless of the specific lighting (passive vision).

Techniques of active vision

You are techniques are what make speak a specific light source to determine the three-dimensional coordinates of the points of measurement. Optical systems are based on the calculation of depth.

These optical systems are always at least a light transmitter and a receiver. Knowing the direction of the emitted beam and received the dimensions of the formed triangle are obtained and therefore to obtain the depth of the inspected item.

There are three types of optical systems:

2.1 Scanning laser

The source of light in these systems is a diode laser. This laser is planning a line of light on the surface we are going to digitize. The reflected light will be detected by one or two photosensitive cells that are located on both sides of the laser. These detectors read the beam of light reflected and transmitted information on the profile of the piece to a software. The result of scanning gives us the possibility of a reverse engineering, i.e., obtain the full geometry of the workpiece irrespective of the levels.

2.2 Structured light projection

In this system the emitter is a white light projector and the receiver a CCD camera.

When starts scanning the projector launches a series of vertical light stripes of light on the subject and shadows alternate, that they are recorded by the camera. To obtain the coordinates of points scanned by triangulation, it could mathematically said that projector system is a transmitter of planes of light and the CCD camera a receiver of straight lines. Calculation of depth is to solve the plano-recta intersections.

2.3 Telemetry

Telemetry is to measure the time of route of a (laser) light beam to the surface of measure. It can be measured in two ways (with the measurement of time of flight or the calculation by phase difference). In the first case the data are obtained by measuring the time between the emission of the light impulse and the observation of the return. The second regulates the luminous momentum following a frequency of finished and the gap between the emitted beam and returnee light is measured.

The process of reverse engineering

AIJU has acquired a digitalized 3D Minolta VI - 910 laser and RapidForm software for the treatment of the cloud of points. This digitizer with the process of reverse engineering would be composed of the following stages:

(a) digitalization of the piece

Definition of parameters of work.
Selection of the appropriate lens, depending on the size and geometry of the workpiece and the required precision.
Adjust the power of the laser to beam to reflect properly on the surface of the piece.
Placement of the piece.

With the 3D scanner carried out sweeps of the object from different points of view. Each scan gives rise to a map of depth (range map) of the visible area to the apparatus, and associated with each point captured with the color information.

These maps of depth are treated to eliminate areas with defects or elements unrelated to the piece.

Different maps are transformed into meshes of triangles (mesh).

Align different meshes and docks for one model of 3D triangles.

Editing operations are needed to correct small errors, simplify redundant data and smooth the mesh.

Optimized once the mesh can be exported to an STL file (for the realization of a prototype or continue with the process of reverse engineering).

(b) creation of surfaces

The RapidForm allows for NURBS surfaces from the mesh of triangles 3D.

The way of doing this is through the creation of four curves on the mesh to be a closed area, and the subsequent construction of the NURBS surface defined within it and approaching such a mesh.

Obtained surfaces must be more like possible triangles 3D mesh, this is carried out a verification of the error as you approach the surface model into triangles.

With the acquisition of the laser digitizer and the processing software point clouds, along with the head of palpation in continuous Renishaw SP 600, AIJU provides a range of new services that are intended to solve the problems of the companies moldistas, paste-up and processing of plastics, providing a complete and perfect service that adapts to the technological needs of the market. For this reason, digitization services include:

Obtaining of the model in mesh polygons for the realization of prototypes, 3D catalogs, etc.
Extraction of curves.
To obtain the model surfaces.
Reverse engineering of the model.
Dimensional control of the piece by comparison with CAD file.