5m), for construction mapping (up to 10m), or for ITS applications (20m). On the other hand, laser triangulation is a good choice for applications with fast-moving parts, at a small distance range (less than 1m), and requiring very high accuracy (down to micrometres, or even lower). For example, electronics inspection, wood inspection or automotive inspection. ‘In the end, each 3D technology has
its own pros and cons, and some of them are more suitable than others for different applications.’ Svorad Stolc, Photoneo: ‘One of the
basic factors that define the achievable acquisition speed is the amount of light. Te amount of light the sensor can use depends on multiple factors, including: the power of the light source (in the case of Photoneo 3D vision systems it is a laser that has specific
‘Te first thing to consider is if your application has static or moving parts and how accurately you need to measure them’
advantages over other illumination types); the working distance (effectiveness of the light emitted by a vision system decreases rapidly with distance from the sensor); parameters of the optics, such as the aperture number, but also the amount of ambient light and other aspects.’
How best do you deal with changing parts or surfaces? Stolc: ‘In general, the best way to deal with diverse scenes or material types (such as matt, glossy, bright or dark materials) is to set up and control environmental conditions in such a way that the sensor works well – which means that its operational dynamic range matches the expected range of the object’s properties. Environmental conditions may include the amount of ambient light, the optimal perspective with regard to the scene or object geometries, and other aspects. Alternatively, one can opt for sensors that are more robust against diverse working conditions. For instance, Photoneo achieves a high scanning robustness of its sensors by a well-rounded hardware and software design featuring laser illumination, optical filters and algorithms for data post-processing. ‘Te robustness of 3D vision systems
may be pushed even further – beyond the possibilities of hardware – through AI, the deployment of which is becoming increasingly popular in 3D vision. Tese
www.imveurope.com | @imveurope
approaches are called data-driven as they leverage the information contained in a training data set describing a given problem domain.’ Nilsson: ‘Dealing with changing or mixed
parts with both very dark and very bright surfaces in the same scene is best done by high dynamic range (HDR) imaging. In our Ruler3000, we use an adjustable, non-linear imager response function to achieve this. Another way to cope with this is to run the sensor in a dual exposure mode – that is, applying one short and one longer exposure time during the same measurement session.’
How complex is 3D vision? How do you reduce complexity? Nilsson: ‘3D vision is not necessarily more complex than 2D vision. With a snapshot 3D camera, such as time-of-flight or stereo cameras –for example, our Visionary cameras – you get a 3D image directly and can apply suitable software tools, just as in 2D vision. Furthermore, it is very common to do the image processing on rectified 3D depth-map images, rather than on a 3D point cloud. In this way, the 3D data can, in many cases, be processed with the same tools as for 2D images, which to a large extent removes the complexity issues of image processing. ‘When using laser triangulation, you do
need to take the movement of parts into consideration, which slightly increases the complexity. However, once set up, the application can be solved by a point-and- click approach, as in our TriSpector1000 series, which does not require a high level of machine vision competence. Actually, for a range of applications, the solution can be even more simple in 3D vision compared to 2D, as illumination is most often included in the device and you get factory calibrated data (such as in Ruler3000). Tis is a clear benefit as parts will remain the same size regardless of where in the field of view they are presented.’ But Lochardet said: ‘Tinking of imager
integration, 3D vision is much more complex than a conventional 2D vision. It involves optics and an illumination system that
g
Power up your images!
4
127 MP @ 17 fps
The new shr661. A combination of global shutter and highest resolution in perfection for ultimate imaging results.
> Global shutter > 13,392 x 9,528 pixels resolution > Smallest hi-res camera available (80 x 80 x X mm) > Industry-leading integrated feature set
Tyre analysis is one of the applications that can be tackled with triangulation
SVS-Vistek | Germany
www.svs-vistek.com
Teledyne Dalsa
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
