caa a22 4500 445295406 CHVBK 20180317142540.0 cr unu---uuuuu 170323e20101201xx s 000 0 eng 10.1007/s11554-010-0159-4 doi (NATIONALLICENCE)springer-10.1007/s11554-010-0159-4 A two-level real-time vision machine combining coarse- and fine-grained parallelism [Elektronische Daten] [Lars Jensen, Anders Kjær-Nielsen, Karl Pauwels, Jeppe Jessen, Marc Van Hulle, Norbert Krüger] In this paper, we describe a real-time vision machine having a stereo camera as input generating visual information on two different levels of abstraction. The system provides visual low-level and mid-level information in terms of dense stereo and optical flow, egomotion, indicating areas with independently moving objects as well as a condensed geometric description of the scene. The system operates at more than 20Hz using a hybrid architecture consisting of one dual-GPU card and one quad-core CPU. The different processing stages of visual information have rather different characteristics that in some cases make fine-grained parallelization on a GPU less applicable. However, for most of the stages that are not efficiently implementable on a GPU, a coarse parallelization on multiple CPU-cores is applicable. We show that with such hybrid parallelism, we can achieve a speed up of approximately a factor 90 and a reduction of latency of a factor 26 compared to processing on a single CPU-core. Since the vision machine provides generic visual information it can be used in many contexts. Currently it is used in a driver assistance context as well as in two robotic applications. Springer-Verlag, 2010 Hybrid parallelism nationallicence Early vision nationallicence Mid-level vision nationallicence GPU computing nationallicence Multi-core nationallicence Jensen Lars The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut Kjær-Nielsen Anders The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut Pauwels Karl Laboratorium voor Neuro- en Psychofysiologie, K.U.Leuven, O&N 2 Herestraat 49, bus 1021, 3000, Leuven, Belgium aut Jessen Jeppe The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut Van Hulle Marc Laboratorium voor Neuro- en Psychofysiologie, K.U.Leuven, O&N 2 Herestraat 49, bus 1021, 3000, Leuven, Belgium aut Krüger Norbert The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut Journal of Real-Time Image Processing Springer-Verlag 5/4(2010-12-01), 291-304 1861-8200 5:4<291 2010 5 11554 https://doi.org/10.1007/s11554-010-0159-4 text/html Onlinezugriff via DOI 1 research-article jats NATIONALLICENCE 856 40 https://doi.org/10.1007/s11554-010-0159-4 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Jensen Lars The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut NATIONALLICENCE 700 1- Kjær-Nielsen Anders The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut NATIONALLICENCE 700 1- Pauwels Karl Laboratorium voor Neuro- en Psychofysiologie, K.U.Leuven, O&N 2 Herestraat 49, bus 1021, 3000, Leuven, Belgium aut NATIONALLICENCE 700 1- Jessen Jeppe The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut NATIONALLICENCE 700 1- Van Hulle Marc Laboratorium voor Neuro- en Psychofysiologie, K.U.Leuven, O&N 2 Herestraat 49, bus 1021, 3000, Leuven, Belgium aut NATIONALLICENCE 700 1- Krüger Norbert The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Campusvej 55, 5230, Odense M, Denmark aut NATIONALLICENCE 773 0- Journal of Real-Time Image Processing Springer-Verlag 5/4(2010-12-01), 291-304 1861-8200 5:4<291 2010 5 11554 Metadata rights reserved Springer special CC-BY-NC licence nationallicence BK010053 XK010053 XK010000 NATIONALLICENCE NATIONALLICENCE NL-springer