Time Delay Integration - Roper Photometric Owner's Manual

1
The CCD is exposed
to light and a charge
pattern accumulates
in the parallel register.
5
The summed charge
is shifted again. The
output node contains
the charge from four pixels.

Time Delay Integration

34 Advanced Camera Operation Manual
Binned Readout
2
Charge in the parallel
register is shifted once
into the serial register.
6
The charge at the output
node is collected
for signal processing.
Time delay integration (TDI) is an integration and readout mode that allows the
acquisition of long swaths of a moving image.
The figure Time Delay Integration on page 35 illustrates the TDI principle. A
moving image is focused on an unshuttered CCD imager. The parallel register is
clocked in step with image motion, so that charge packets always correspond to
the same image region as they move across the parallel register. Charge
accumulates and signal strength increases as the pixels approach the serial
register. When pixels reach the serial register, they are transferred out, digitized,
and stored in the normal fashion. The exposure time for each pixel is exactly the
length of a full parallel shift sequence, which is determined by the velocity of the
scene.
Compared to a simple exposure, TDI increases sensitivity in proportion to the
number of rows in the parallel register. Although the TDI technique was first
used by astronomers to take images of passing star fields, it has also been used
to acquire swath images from airplanes and satellites, and is very effective for
inspecting articles on a moving conveyor belt. TDI is normally used with scenes
moving at a constant velocity, but other variations are possible.
3
Charge is again shifted
into the serial register.
The serial register con-
tains the summed
charge.
7
Steps 4 and 6 are
repeated until the entire
serial register is read out.
4
The summed charge is
shifted once into the
output node.
8
Steps 2 through 7
are repeated until
the entire parallel
register is read out.