PMD sensor

from Wikipedia, the free encyclopedia

A photonic mixer device , including PMD sensor called Photonic Mixing Device , is an optical sensor whose principle of operation on the light transit time method is based (ger .: Time of Flight) and is often used as a synonym for all TOF sensors. A PMD sensor is often used as an image sensor in TOF cameras .

history

The research team for the PMD chip designed by Rudolf Schwarte in 1996 was nominated for the German Future Prize in 2002. A PMD 3D camera from S-Tec, today Pmdtechnologies, was already available in 2000 . From 2008 to 2012 the humanoid robot Justin of the Institute for Robotics and Mechatronics ( DLR ) was equipped with PMD sensors.

Working principle

The measuring objects are illuminated by light pulses and the signal transit time is measured. The distance between the camera and the object can be calculated based on the running time. This semiconductor - component enables distances directly observed. The resulting distance image can then be displayed in various ways (e.g. colors as distances). In addition to the distance, a gray-scale image can be calculated from the intensity of the reflected light.

The modulated light signal transmitted by a transmitter, e.g. B. invisible infrared light illuminates the scene to be measured. The light reflected from the scene hits the PMD sensor. This is also coupled to the modulation source. Thus, the converted photons into electrons in dependence be in the photosensitive semiconductor region of the reference signal pixel by pixel, separated by means of so-called charge carrier rocking distance selectively.

Through this simple comparison process between the optical measurement and the electronic reference signal, the resulting output signal of the sensor already represents a direct reference to the 3D information. A particular advantage of the PMD system is that an efficient suppression of extraneous light (e.g. Solar radiation) is achieved. The active transmitter signal is filtered out of the ambient light and thus enables use even under difficult ambient conditions.

  • Examples of PMD cameras

  • "PMDs CamCube with 204 × 204 pixels and SBI"

  • "SwissRanger 4000 from MESA Imaging with 176 × 144 pixels"

  • "TOF camera from the European ARTTS project"

Ambiguities

A disadvantage of PMD sensors, as with all other distance measurement methods that use the phase difference method, is that an unambiguous determination of the distance is not always possible. Since the distance is determined indirectly by the phase shift between the transmitted and received signal, ambiguities can arise, i.e. H. the measured distance could also be the measured distance plus a multiple of the wavelength.

A possible solution to this ambiguity problem is the use of several wavelengths or specific code words that allow a range of uniqueness well over 500 m.

Suppression of extraneous light

In addition to the properties described above, PMD sensors have the ability to actively suppress extraneous light. Due to the correlation property of the sensor, uncorrelated parts of the incident light can be deducted from the correlated part of the light directly in the pixel via a special circuit, the so-called SBI ( Suppression of Background Illumination ). The dynamic of the pixel is then completely available to the active light. The PMD sensor can be operated with the SBI circuit in full sunlight of 150  klx .

literature

  • Bernd Buxbaum: Optical transit time measurement and CDMA based on PMD technology using phase-variable PN modulation . Shaker Verlag , Aachen 2002, ISBN 978-3-8265-9805-0 .
  • Wei Tai: Investigations of 3D PMD cameras with special consideration of optical optimization . Shaker Verlag, Aachen 2001, ISBN 978-3-8265-8789-4 .
  • R. Schwarte , H. Heinol, B. Buxbaum, Z. Xu, T. Ringbeck, Z. Zhang, W. Tai, K. Hartmann, W. Kleuver, X. Luan: Novel 3D vision systems based on layout-optimized PMD structures . In: tm - technical measurement . No. 7-8 , 1998, pp. 264-271 .

Web links

Individual evidence

  1. ↑ Christoph Heckenkamp: The magic eye - Basics of image processing: The PMD principle . In: Inspect. No. 1, 2008, pp. 25-28.
  2. ↑ Norbert Lossau : 3D camera records your spatial environment in real time . In: Die Welt , December 3, 2002.
  3. ↑ Ling Shao, Jungong Han, Pushmeet Kohli, Zhengyou Zhang (eds.): Computer Vision and Machine Learning with RGB-D Sensors . Springer Science + Business Media , 2014, ISBN 3319086502 , p. 15.
  4. ↑ Automatica 2008 - Mobile humanoid upper body Justin , Institute for Robotics and Mechatronics (DLR)
  5. ↑ Mechatronic design by Rollin 'Justin , Institute for Robotics and Mechatronics (DLR), as of 2016, page has now changed.