Share Email Print

Proceedings Paper

Adaptive optical flow detection using correlation image sensor and frequency-tuned complex-sinusoidal reference signals
Author(s): D. Wei; P. Masurel; T. Kurihara; S. Ando
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The optical flow computation has been widely studied motivated by a broad range of applications. We proposed a solution base on optical identity (OFI) using Correlation image sensor (CIS). If the frequency of the CIS' reference signals is ill-chosen, the SNR of the amplitude of the complex correlation image is small, and it is difficult to solve OFI stably. When we choose the frequency of the sinusoidal reference signals "w" satisfies "wT=2n Pi" the optical flow identity (OFI) holds, where "n" is an integer, "T" is the frame time of the CIS, and "Pi" is the circle ratio. We define "G(w)" satisfying "wT=2nPi" by the sum of the amplitude of the complex sinusoidally modulated image over all pixels, and maximize it to give a solution of the frequency-tuned problem.

Paper Details

Date Published: 29 February 2008
PDF: 11 pages
Proc. SPIE 6816, Sensors, Cameras, and Systems for Industrial/Scientific Applications IX, 68160M (29 February 2008); doi: 10.1117/12.767141
Show Author Affiliations
D. Wei, The Univ. of Tokyo (Japan)
P. Masurel, The Univ. of Tokyo (Japan)
T. Kurihara, The Univ. of Tokyo (Japan)
S. Ando, The Univ. of Tokyo (Japan)

Published in SPIE Proceedings Vol. 6816:
Sensors, Cameras, and Systems for Industrial/Scientific Applications IX
Morley M. Blouke; Erik Bodegom, Editor(s)

© SPIE. Terms of Use
Back to Top