A vision system for 2D real-time vector motion estimation of a random-sized moving pattern

A 2D motion detection and estimation vision system is here presented. It performs real-time template matching on edges extracted from successive frames, acquired with an addressable CMOS camera. The x-y motion vector components are estimated counting the number of matched edges moving over less than one pixel within two subsequent frames. The algorithm has been implemented on an FPGA and uses a 20 times 20 pixel binary- image. The system computes up to 350 motion-vector-estimation/s on a random-sized pattern. The computation rate is mainly limited by the frame rate of the camera. The system was linked to a PC through an USB interface.

[1]  David Stoppa,et al.  A CMOS image sensor with programmable pixel-level analog processing , 2005, IEEE Transactions on Neural Networks.

[2]  Javier Díaz,et al.  FPGA-based real-time optical-flow system , 2006, IEEE Transactions on Circuits and Systems for Video Technology.

[3]  P. Martinet,et al.  The task "template tracking" in a sensor dedicated to active vision , 2003, 2003 IEEE International Workshop on Computer Architectures for Machine Perception.

[4]  Chung-Yu Wu,et al.  A CMOS focal-plane motion sensor with BJT-based retinal smoothing network and modified correlation-based algorithm , 2002 .

[5]  Guido Torelli,et al.  A single-chip optical sensor with analog memory for motion detection , 1995, IEEE J. Solid State Circuits.

[6]  X. Arreguit,et al.  A CMOS motion detector system for pointing devices , 1996, 1996 IEEE International Solid-State Circuits Conference. Digest of TEchnical Papers, ISSCC.

[7]  Jörg Kramer.  Compact Integrated Motion Sensor With Three-Pixel Interaction , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Idaku Ishii,et al.  A digital vision chip specialized for high-speed target tracking , 2003 .

[9]  Massimo Gottardi,et al.  A CCD/CMOS image motion sensor , 1993, 1993 IEEE International Solid-State Circuits Conference Digest of Technical Papers.