STH-DCSG
U
SER
’
S
M
ANUAL
©
2005
V
IDERE
D
ESIGN
20
the light on the image will be adjusted by changing the exposure.
For high-speed motion, it is better to set the exposure to a fixed, low value,
and then use auto-gain to keep reasonable light levels on the imager.
7.4 Subsampling
In many applications it is not necessary to work with the full 640x480
image. The CMOS imagers are capable of sampling the pixels in the array.
Sampling allows the video stream to send less data, for faster frame rates or
less bus activity. A sampled image shows the same scene as the original
image, but it uses fewer pixels to do so, and has less detail.
Binning
is a subsampling technique in which several adjacent pixels are
averaged into one. Binning reduces video noise, sometimes quite
substantially. Binning is available on-chip (monochrome only) and on the
host PC – the larger image is first transferred on the IEEE 1394 bus, then
binned down.
Sampling differs from
subwindowing
, which picks a rectangular portion of
the image, but doesn’t change its resolution. The STH-DCSG has one
subwindow mode, in which the center 320x240 subwindow of the imager is
chosen. Using the subwindow has the effect of zooming the image by a
factor of 2.
Figure 7-2 shows the frame size and subsampling controls on the video
capture application. With SVS version 3.2b, the sampling control has been
changed to a simple Frame Division control. For the STH-DCSG, there are
two possible values, 1 (full image) and 1/2.
Refer to Table 4 for a complete list of allowed modes, and how the frame
size and sampling setting affect the output image. Explicit control over the
sampling mode is accomplished with the
SetSample()
function from the
SVS API.
The first two lines of the table are for full-frame images. At 640x480, the
full image is sent to the host PC, and there is no binning. At 320x240, there
are two choices: binning on the imager (monochrome only), or binning on
the host; the default for monochrome is binning on the imager, and for
color, binning on the host PC. Using
SetSample(2,1)
means that
binning will take place on the imager, which then transmits a 320x240
imager. Using
SetSample(1,2)
means that binning will take place on
the host. In this case, the imager transmits all 320x240 pixels, so the
maximum frame rate is lower. This is the greyed-out line in the table.
For 1/2 frame images, the resolution is 320x240. In this case, the STH-
DCSG transmits a 320x240 image.
7.5 Frame
Rates
Frame rates from the STH-DCSG/-C depend on the frame size. Table 4
shows the frame rates available for each of the frame sizes. Note that a 50
Hz option is available – see Section 3.3
7.6 Firmware
Parameters
There is one firmware parameter that affects the overall behavior of the
STH-DCSG.
•
50 Hz operation
This parameter can be changed by using the Firmware Parameter dialog,
accessible from the
smallv
menubar. Choosing this menu brings up the
dialog, which is shown in Figure 7-3.
The dialog lists many of the internal parameters of the device, which are
Resolution Frame
Bin
on
imager
Bin on
PC
Frames per Second
640 x 480
Full
no
no
3.75, 7.5, 15, 30
Monochrome only
320 x 240
Full
yes
no
3.75, 7.5, 15, 30, 60
320 x 240
Full
no
yes
3.75, 7.5, 15, 30
320 x 240
1/2
no
no
3.75, 7.5, 15, 30, 60
Table 4 Subsampling modes and frame rates for the STH-DCSG.