Standards
@imveurope
www.imveurope.com
GenICam fit for embedded and 3D vision
Newly released GenICam 3.0, hosted by the European Machine Vision Association, includes, among other functionality, standardised connection of 3D cameras. Andreas Breyer, EMVA director of public relations and market research, reports on the upgrade
I
maging in 3D has become very popular in machine vision; a multitude of 3D cameras are now available from different vendors. Terefore,
for standardised connection of 3D cameras, manufacturers and soſtware developers have developed a uniformly accepted standard which not only offers much easier integration but also enables inexperienced users easy access to 3D machine vision. Te most popular 3D camera technologies
are sheet of light and structured light methods, although cameras based on time-of-flight and stereovision are also available. Still, most integrators tend to rely on standard 2D cameras at the moment due to the lack of compatibility between machine vision soſtware and 3D cameras. Te different 3D technologies also bring different implementations of the various camera manufacturers, and, in many cases, the camera maker requires a special driver or proprietary programming interface (API) to operate each respective camera. Tis problem has been addressed and solved for 2D cameras already through the GenICam standard, which made a significant contribution to the success of machine vision, and now takes this success to the next level through the integration of 3D cameras. In the GenICam 3.0 release access to 3D cameras
is now standardised. Terefore, communication and image acquisition between the application (soſtware) and the camera takes place on a uniformed basis and enables a controlled exchange which applies across various camera manufacturers and interfaces (GigE Vision, USB3 Vision, CoaXPress, and Camera Link HS). Furthermore, proprietary transport layers can make use of the new standardised GenICam 3D interface. ‘Since practically all soſtware solutions support
the standard, GenICam-compliant cameras are being recognised and put into operation automatically by the soſtware,’ said Stephan Kieneke, 3D product manager at Automation Technology and member of the GenICam standard committee. ‘Tis facilitates the handling in particular for users
with little experience in 3D imaging since they can access the 3D camera in a familiar working environment.’ For instance, under the new 3D functions of
GenICam integrators will soon be able to extract 3D point clouds out of the cameras and process them directly in their soſtware without circumstantial data conversion. ‘Tis will make it possible to perform the transformation of 2D pixels to 3D world coordinates either within the camera or have it calculated by the soſtware according to the given standardised transformation parameters,’ explained Christoph Zierl, technical director at MVTec Soſtware and vice-chair of the GenICam standard committee. ‘Until now, this process was only possible through a proprietary manufacturer setting which complicated usability and interchangeability for the customer.’ A problem which has been solved with the introduction of GenICam 3.0, since no matter which technology hides behind a 3D camera, the user always receives a standardised data format. Te expansion of GenICam
for 3D cameras means that the dissemination of 3D applications could increase considerably and 3D be established as future oriented solutions in many industries. Machine vision users from other industries now have easy access to 3D image processing which sets the scene for a much broader usage of this technology in the coming years. Te 3D extensions are already available for new GenICam module versions of SFNC, PFNC, and GenTL, and the respective documents can be downloaded at
www.genicam.org.
GenAPI for embedded vision In a separate project GenAPI is currently under revision to make it fit for the future. GenAPI contains the reference implementation of the GenICam standard or in other words the
28 Imaging and Machine Vision Europe • Yearbook 2015/2016
programming code which allows a machine vision application generic access on all features of any GenICam compatible device. Technically this means that the XML file with all camera features is being read out and all features are accessible by means of a generic API. GenAPI is available for Windows 32-bit, Windows 64-bit, Linux, OS X and Linux on Arm processors, and it is used as it is by a lot of manufacturers of cameras, frame grabbers, and vision soſtware. Zierl explained the need for the fundamental
Machine vision users from other industries now have easy access to 3D image processing
revision of GenAPI that’s currently being undertaken: ‘Te basic soſtware architecture for GenICam was chosen around 2005. At that time, the PC was the target platform of practically all applications and camera description XML files were assumed to be small. Today, 10 years later, things have changed. Embedded, oſten even deep embedded systems, are now used for image processing. In
addition, today’s cameras are extremely feature- rich and bear huge XML files. By this GenICam v2.4 oſten is too slow and too fat for embedded systems. Tis is why the standard committee decided three years ago to completely re-write the GenAPI reference implementation from scratch in order to make GenICam faster, smaller, and easier to install. Of course, existing XML files are further supported to maintain backwards compatibility.’ So, the revision should push performance and reduce storage space. GenAPI will see a release candidate in the third quarter of this year; the release of GenAPI 3.0 is planned for the end of 2015. O
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48