Image Processing Reference
In-Depth Information
Image-Stabilization Systems Simplified
51
5.3 System Control
The optical layout shown in Fig. 5.1 requires an additional component to function:
a controller that takes information from the optical sensor and converts the signal
into a correction to the position of the tip-tilt mirror. As such, the controller must be
able to “read in” the voltages from the sensor that define the position of the star on
the sensor, and write to the tip-tilt mirror the movement that keep the star at the de-
sired position. This requires the control system to take the detected motion from the
sensor and convert it to the correct amount of motion in the compensator.
The controller's ability to function correctly requires good calibration of the
range of motion of the compensation device and how that range maps onto the ac-
tive area of the sensor. This calibration map can be created by using the tip-tilt mir-
ror to move a spot of light over the sensor, recording the voltages on the mirror and
the voltages from the sensor. Ideally, the result is a linear relationship between the
two devices. When properly calibrated, the result is an equation that maps the posi-
tion of the light on the sensor to the voltage on the mirror (a proportional control
scheme):
B
A
S
V
=
,
(5.1)
x
x
where
V
is the voltage applied to the compensation mirror, and the ratio of
B
to
A
scales the applied voltage to the mirror to the location of the spot on the sensor.
This astronomical image-stabilization system is designed so that the reference
point for stabilization is wherever the light lands on the sensor. This is often better
than defining a specific point, such as the center of the sensor, as the reference, be-
cause it requires less effort to set up the system. It is important to place the mirror in
a neutral position, with an even amount of throw in all directions, so that the maxi-
mum angular throw is available to keep the image on the camera. Similarly, care
must be taken so that the location of the image on the sensor is not too close to the
edge of the sensor or, if there is a large excursion in the position of the star image, it
could fall off the edge of the sensor. To prevent the image of the star from being too
close to the edge of the sensor or, during operation, having the mirror close to its
limits, an automated approach is commonly used to signal the telescope mount to
reposition itself to a more favorable orientation.
With the control system operating, the image-stabilization system detects and
responds to motion of the spot on the sensor, and uses the tip-tilt mirror to drive the
spot back to the reference point. An integrating camera records the motion of the
spot, but since the excursions away from the stabilization position are kept to a min-
imum, most of the time the spot remains on target. As a result, the intensity of the
peak and the resolution of the image are improved. The next section builds on these
concepts to produce a simple working image-stabilization system.
