Page 438 - Introduction to Information Optics

Exercises 423

7.26 P. Purwosumarto and F. T, S. Yu, 1997, "Robustness of Joint-Transform Correlator versus
Vander Lugt Correlator," Opt. Eng., vol 36, 2775.
7.27 G. Lu, et al., 1997, "Implementation of a Nonzero-Order Joint-Transform Correlator by Use
of Phase-Shifting Techniques," Appl. Opt., vol. 36, 470.
7.28 C. T. Li, et al., 1998, "A Nonzero-Order Joint-Transform Correlator," Opt. Eng., vol. 37, 58
7.29 F. T. S Yu et al., 1998, "Performance Capacity of a Simulated Annealing Bipolar Composite
Filter (SABCF)," Opt. Commun., vol. 154, 19.
7.30 F. T. S. Yu and S. Jutamulia, Optical Pattern Recognition, Cambridge University Press.
Cambridge, UK, 1998.
7.31 F. T. S. Yu and S. Yin, Optical Pattern Recognition, in SPIE Milestone Series, vol. 156, SPIE
Opt. Eng. Press, Bellingham, Wash., 1999.
7.32 M. S. ASam, Optical Pattern Recognition Using Joint Transform Correlator, in SPIE Milestone
Series, vol. 157, SIE Opt. Eng. Press, Bellingham, Wash., 1999.

EXERCISES

7.1 Suppose that the resolution of the SLMs is given by 100 lines/mm and
the CCD array detector is given as 20 lines/mm. Referring to the FPD of
Fig. 7.5, compute the space-bandwidth product of the hybrid-optical
processor.
7.2 Assume that the spatial frequency content of the SLM1 is given as 10
lines/mm. Calculate the space-band width requirement for the SLM2 for
a Vander Lugt spatial filter. Assume that the focal length of the transform
lens is 400 mm.
7.3 Assume that K number of input objects are displayed on the input LCTV
panel of a JTC, shown in Fig. 7.6, as given by

/(x, y) = £ f k(x ~ a fc, y — b k),
* = i

where (a fc, fr fc) denotes the locations of the input objects. The LCTV has an
inherent grating structure so that its amplitude transmittance can be
written as

t(x, y) = f(x, y)g(x, >>),

where g(x, y) represents a two-dimensional grating function of the LCTV.
(a) Compute the output complex light distribution.
(b) Identify the output auto- and cross-correlation terms of part (a).
(c) To ensure nonoverlapping distributions, determine the required loca-
tions of the input functions.