VIDEO PROCESSING IN MOTION MODELLING
Transkript
VIDEO PROCESSING IN MOTION MODELLING Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA [email protected], [email protected], [email protected] Institute of Chemical Technology Department of Computing and Control Engineering Digital Signal and Image Processing Research Group http://dsp.vscht.cz VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.1/10 Contents Introduction VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.2/10 Contents Introduction System Description VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.2/10 Contents Introduction System Description Three-Dimensional Object Detection VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.2/10 Contents Introduction System Description Three-Dimensional Object Detection Motion Visualization VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.2/10 Contents Introduction System Description Three-Dimensional Object Detection Motion Visualization Conclusions VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.2/10 Introduction Goals of the project Study of image acquisition using synchronized two camera system and A/D convertors Study of mathematical methods for body localization in the three dimensional space Visualization of the body movement using virtual reality environment VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.3/10 Introduction Goals of the project Study of image acquisition using synchronized two camera system and A/D convertors Study of mathematical methods for body localization in the three dimensional space Visualization of the body movement using virtual reality environment Application Modelling of the body movement Analysis of the object movement using several reference points VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.3/10 System Description Light @i @ Technical details Cameras with color CCD sensor and with resolution 1024x768, 30 fps A Camera B Camera Synchronization 125 µs Connection with computer via IEEE 1394 Direct connection to the MATLAB system and Image Acquisition Toolbox IEEE 1394 ? ? MATLAB & Image Acquisition Tlbx Computer VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.4/10 Three-Dimensional Object Detection Principle of the object localisation (b) K−TH POSITION (a) INITIAL POSITIONING FRONT VIEW FRONT VIEW C [xC(k),zC(k)] b2(k) C [x (1),0] a (k) 2 α2(k) C β2(k) TOP VIEW TOP VIEW C [x (1),y (1)] C C C [xC(k),yC(k)] b (1) 1 α1(1) a1(1) c b (k) 1 β1(1) α1(k) a1(k) c β1(k) A [xA,yA] B [xB,yB] A [xA,yA] B [xB,yB] Camera A Camera B Camera A Camera B VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.5/10 Three-Dimensional Object Detection Principle of the object localisation (b) K−TH POSITION (a) INITIAL POSITIONING FRONT VIEW FRONT VIEW C [xC(k),zC(k)] b2(k) C [x (1),0] a (k) 2 α2(k) C β2(k) TOP VIEW TOP VIEW C [x (1),y (1)] C C C [xC(k),yC(k)] b (1) 1 α1(1) a1(1) c b (k) 1 β1(1) α1(k) a1(k) c β1(k) A [xA,yA] B [xB,yB] A [xA,yA] B [xB,yB] Camera A Camera B Camera A Camera B 2 2 2 α1 (1) = arccos (b1 (1) + c − a1 (1) )/(2 b1 (1) c) 2 2 2 β1 (1) = arccos (a1 (1) + c − b1 (1) )/(2 a1 (1) c) VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.5/10 Three-Dimensional Object Detection Calibration of the camera system HORIZONTAL AND VERTICAL CAMERA ANGLE EVALUATION (b) INITIAL LIGHT POSITIONING (a) CALIBRATION GRID α2max α (1) 2 d svertical/2 CAMERA α 2min d shorizontal/2 α 1min α1(1) α 1max VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.6/10 Three-Dimensional Object Detection Calibration of the camera system HORIZONTAL AND VERTICAL CAMERA ANGLE EVALUATION (b) INITIAL LIGHT POSITIONING (a) CALIBRATION GRID α2max α (1) 2 d svertical/2 CAMERA α 2min d shorizontal/2 α 1min α1(1) α 1max αhorizontal αvertical = 2 arctan shorizontal /2/d = 2 arctan svertical /2/d VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.6/10 Motion Visualization VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.7/10 Motion Visualization VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.7/10 Motion Visualization MOTION MODELLING 19 30 20 8 z−axis 25 20 18 15 21 2728 7119 10 23 22 17 1 10 16 5 3 15 2 0 26 24 25 4 29 1230 6 CAMERA B −5 13 5 14 1500 1500 CAMERA A 1000 1000 500 y−axis 500 0 0 x−axis VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.7/10 Motion Visualization MOTION MODELLING 19 30 20 8 z−axis 25 20 18 15 21 2728 7119 10 23 22 17 1 10 16 5 3 15 2 0 26 24 25 4 0 29 1230 6 CAMERA B −5 13 5 Display 14 1500 1500 CAMERA A 1000 1000 500 y−axis 500 0 simin 0 Bally.translation x−axis From Workspace VR Sink Scope VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.7/10 Motion Visualization MOTION MODELLING 19 30 20 8 z−axis 25 20 18 15 21 2728 7119 10 23 22 17 1 10 16 5 3 15 2 0 26 24 25 4 0 29 1230 6 CAMERA B −5 13 5 Display 14 1500 1500 CAMERA A 1000 1000 500 y−axis 500 0 simin 0 Bally.translation x−axis From Workspace VR Sink Scope VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.7/10 Conclusions Results VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 Conclusions Results Successfully tested system with one moving object VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 Conclusions Results Successfully tested system with one moving object Creation of the virtual reality model based on the real movement VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 Conclusions Results Successfully tested system with one moving object Creation of the virtual reality model based on the real movement Further Research VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 Conclusions Results Successfully tested system with one moving object Creation of the virtual reality model based on the real movement Further Research Deterministic and statistical analysis of the set of moving objects VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 Conclusions Results Successfully tested system with one moving object Creation of the virtual reality model based on the real movement Further Research Deterministic and statistical analysis of the set of moving objects Their proper recognition and detection followed by visualization in specific applications VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.8/10 References 1. R. Boulic, P. Fua, L. Herda, M. Silaghi, J.S. Monzani, L. Nedel, and D. Thalmann.An Anatomic Human Body for Motion Capture. In Technologies for the Information Society: Developments and Opportunities. EMMSEC98, 1998. 2. J. Lasenby and A. Stevenson. Using Geometric Algebra for Optical Motion Capture. In E.Bayro-Corrochano and G. Sobcyzk, editors, Applied Clifford Algebras in Computer Science and Engineering. Birkhauser, Boston, U.S.A., 2000. 3. M. Kubíček. Using Dragonfly IEEE-1394 Digital Camera and Image Acquisition Toolbox. In Sborník konference MATLAB 2004, pages 280–282. VŠCHT Praha, 2004. 4. M. Nixon and A. Aguado. Feature Extraction & Image Processing. NewNes Elsevier, 2004. 5. M. Ringer, T. Drummond, and J. Lasenby. Using Occlusions to Aid Position Estimation for Visual Motion Capture. In Proc Computer Vision and Pattern Recoginition (CVPR). IEEE USA, 2001. 6. M. Ringer and J. Lasenby. Modelling and Tracking of Articulated Motion from Multiple Camera Views. In Proc. British Machine Vision Conf (BMVC), pages 172–181, 2000. VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.9/10 Thank You! http://dsp.vscht.cz VIDEO PROCESSING IN MOTION MODELLING – Miroslav KUBÍČEK, Aleš PROCHÁZKA, Aleš PAVELKA – Process Control 2005 – p.10/10
Podobné dokumenty
Následky dopravních nehod seniorů na jízdních
2. Český statistický úřad. Populační prognóza ČR do r. 2050. 1. vyd.
2009. http://www.czso.cz/csu/2004edicniplan.nsf/p/4025–04.
3. Český statistický úřad. Vnější příčiny úmrtí v ČR v letech 1994 až...
Beskyd Cup 2013
H10C, H12C, H14C, H16C, H18C, H20, H21E, H21B, H21C, H35C, H40, H45C, H50, H55C, H60,
H65C, H70.
D10C, D12C, D14C, D16C, D18C, D20 (jen sobotní závod), D21E, D21B, D21C, D35C, D40, D45C,
D50, D55C,...
POKYNY Velkobystřické jarní kufříky
Závodníci kategorie HDR obdrţí pamlsek při vyčítání.
Protesty:
POSLEDNÍ KUFR 2015
D10C, D12C, D14C, D16C, D18C, D21C, D35C, D45C, D55C,
H10C, H12C, H14C, H16C, H18C, H21L, H21K, H35C, H45C, H55C, H65C,
DH10N – smajlíci (bez doprovodu), T – tréninková trať (4-6km)
P – trať pro ...
26 COMMUTATING CAPACITORS (only for spare parts
see table / viz. tabulka
-40 / +85 °C
max 85 °C
120 000 h