Biomedical Engineering Reference
In-Depth Information
Figure 3.15b shows the time-integrating system. “Here the integration is
performed in time (rather than space) on the output detector. The output cor-
relation appears as a function of distance across the output detector array.
An input light source such as an LED or a laser can be modulated with the
received signal g ( t ). Lens L 1 collimates the output, and an AO cell at P 1 is uni-
formly illuminated with the time-varying light distribution g ( t ). The trans-
mittance of the cell is now described by h ( t − τ), where again τ = x/v . The light
distribution leaving the cell is g ( t ) h (t − τ). Lenses L 1 and L 2 image P 1 onto P 3
(with SSB filtering performed at P 2 ), where time integration on a linear PD
array occurs. The P 3 light distribution (after time integration occurs on the
detector) is thus
U x
3 ( )
=
h t
(
τ
)
g t
(
τ
)
d
t
=
g
h
(3.14)
or again the correlation of the received and reference signals. In this case, the
integration is performed in time and the correlation is displayed in space.
The optical correlators of Figures 3.15a and b employ 1-D devices and sim-
ple imaging lenses and are relatively easy to implement. They realize the
correlation operation with a moving-window transducer without the need
for a matched spatial filter as in conventional correlation optical processors
with fixed-format transducers. The space-integrating system can accommo-
date large range-delay searches between the received and reference signals;
however, the signal integration time and time-bandwidth product that this
system can handle are small, limited by the aperture (40 μs dwell time is
typical) and TBWP (1000 is typical) of the AO cell.
In the time-integrating system, the signals must be time aligned, and only
a much smaller range-delay search window (equal to 40 μs typical aperture
time of the cell) is possible; however, the time-integrating processor allows
longer integration time (limited by the integration time and noise level of the
detector) and the associated correlation of longer TBWP signals” [32].
3.10 OpticalLogicGates
The AO signal processing systems discussed previously may be consid-
ered representative of the traditional analog optical computing technology.
Technological niches for analog optical computing include image enhance-
ment and noise reduction, spectrum analysis of RF signals, pattern recogni-
tion, and signal correlation in radar, sonar, and guidance systems. Another
class of devices is directed toward the development of digital optical com-
puter systems. Combinations of only 10-20 of these gates could potentially
 
Search WWH ::




Custom Search