New 3D sensor scans transparent objects
A new measurement method for 3D shape acquisition has been developed by researchers at the Fraunhofer Institute for Applied Optics and Precision Engineering IOF. With their “MWIR 3D sensor”, they can scan objects three-dimensionally, regardless of whether they are made of transparent plastic or glass. Even objects with shiny metallic or jet-black surfaces can be detected without any difficulties. Combining different materials is also no problem for the new 3D infrared sensor. In the field of 3D sensor technology, this degree of flexibility in the properties of the object is a first. Applications are conceivable in areas such as quality control in production and robotics.
Until now, if you wanted to measure reflective, transparent or black surfaces accurately
with conventional 3D scanners, you first had to treat the surface for this purpose. This
means that the objects were temporarily coated with varnish for the measurement.
After the scan, this usually had to be removed again at great effort.
Glass and transparent objects become visible to machines
The latest invention by Fraunhofer IOF will make this impractical and time-consuming
treatment of the object superfluous in the future. Due to the size of the measuring
field as well as the resolution and speed, the method is also suitable for quality control
in production processes or for applications in automation.
This is possible because researchers at the Fraunhofer Institute in Jena, Germany have
succeeded in making thermal radiation usable for 3D measurement. The researchers
therefore refer to this method as “3D sensing in the thermal infrared range”. At the
heart of the system is a high-energy CO2 laser with which the objects are irradiated.
Using special lenses for high power densities, the laser beam is expanded into a line
that vertically illuminates the entire object. For a high-resolution measurement result,
this line is moved over the object in a specially coordinated sequence. The energy of
the laser light is absorbed by the measured object and partially re-emitted.
Combination of thermography and triangulation
Two thermal imaging cameras analyze the thermal signature left by the narrow and
intense infrared line on the object from two different perspectives. Afterwards, a software developed in-house calculates spatial pixels from the information of the two viewing angles and merges them into the exact dimensions of the measured object.
The thermal energy introduced for the 3D analysis is so low that the object is not damaged. The temperature difference between heated and non-heated surfaces is typically
less than 3 °C. For this reason, the method is also suitable for sensitive materials.
“By switching from a full-surface thermal pattern to a narrow thermal strip, we have
succeeded in advancing the technology in such a way that we can meet the requirements placed on a 3D sensor in industrial use,” emphasizes Martin Landmann, a researcher of the “Imaging and Sensing” department at Fraunhofer IOF. Together with
his team and a group of researchers of the innovation alliance “3Dsensation”, he has
been working on the system since 2017.
“With adaptive mirror optics, we have succeeded in focusing the power of the laser on
a much smaller surface, thus providing the necessary contrast for the thermal imaging
cameras much faster. Only this made it possible to achieve an accuracy of less than 10
µm for the 3D coordinates with a field of view width of 160 mm,“ he explains.
Conceivable applications in robotics
Following the successful scientific demonstration of the new measurement method, the
researchers are now working intensively to make the measurement principle ready for
the market: “For us, it is now a matter of transferring the system from the laboratory
to practical use,” explains Landmann. He already has concrete areas of application in
mind: “The parameters of our system allow us to optimize it for different application
scenarios. If we reduce the resolution to below 50 µm, we can record a stereoscopic
data set in under a second and are thus fast enough for applications in robotics.”
At Fraunhofer IOF, researchers are currently developing various systems based on the
MWIR 3D measurement method. In addition to optimizing the method for various
measurement scenarios and applying it in industrial plants, the team led by Martin
Landmann and group leader Dr. Stefan Heist is working on a system for use in robotics.
This system focuses on transforming the laboratory setup into a prototype that is as
compact and robust as possible. In this way, robots can be enabled to recognize and
grasp transparent objects.
First MWIR 3D system „Glass360Dgree“ presented at „Control-Virtual“
The first application-oriented system to use this MWIR 3D measurement principle is
„Glass360Dgree“. The system is specifically designed for inspecting glass bodies in
optics manufacturing and will also be used by research partners in the further course of
testing how the measurement method can be integrated into a wide variety of robotic
processes. The Fraunhofer IOF researchers will present „Glass360Dgree“ to the public
for the first time at „Control-Virtual“, the international trade fair for quality assurance,
starting May 3.
Trade journal “inVISION” honors new sensor as “Top Innovation 2021”
The new 3D measurement method in the thermal infrared region was recently selected
as “Top Innovation 2021” by the magazine “inVISION” – a trade journal for the topics
of image processing, embedded vision and measurement technology. “We are very
pleased about the inVISION award,” says group leader Dr. Stefan Heist. “This is a wonderful confirmation of our intensive work over the past years and a great motivation to
keep enhancing our 3D thermal system.”
Dr. Stefan Heist
Imaging and Sensing
Phone: +49 (0) 3641 807-214
https://www.iof.fraunhofer.de/en/pressrelease/2021/3d-mwir-en.html Press release (incl. images and video demonstration) on the website of Fraunhofer IOF
https://www.iof.fraunhofer.de/en/competences/measurement-methods-and-characterization/3d-measurement.html 3D measuring systems at the Fraunhofer IOF
https://www.iof.fraunhofer.de/en/competences/measurement-methods-and-characterization.html Imaging and Sensing at Fraunhofer IOF