Instruction manual
Guppy Technical ManualV7.4.0
110
Description of the data path
Building defect pixel data
To build defect pixel data perform the following steps:
Grab an image with defect pixel data
1. Take the camera, remove lens and put on lens cap.
2. Set image resolution to Format_7 Mode_x or Format_7 Mode_0 (when
using fixed modes) and set AOI to maximum.
3. Set values for shutter, gain and brightness (offset) to maximum.
4. Grab a single image (one-shot).
Calculate defect pixel coordinates
5. Accept default threshold from system or choose own threshold.
6. Set the BuildDPData flag to 1.
In microcontroller the defect pixel calculation is started. The detected
defect pixel coordinates are stored in the dual port RAM of the FPGA.
Defect pixel coordinates are:
– 16-bit y-coordinate and
– 16-bit x-coordinate
Note
• Defect pixel correction is only possible in Mono8 modes
for monochrome cameras and Raw8 modes for color
cameras.
• In all other modes you get an error message in advanced
register 0xF1000298 bit [1] see Table 115: Advanced reg-
ister: Defect pixel correction on page 243.
• Using Format_7 Mode_x: Defect pixel correction is done
in Format_7 Mode_x.
• Using a fixed format (Format_0, Format_1 or Format_2):
Defect pixel correction is done in Format_7 Mode_0.
• When using defect pixel correction with binning and
sub-sampling: first switch to binning/sub-sampling
mode and then apply defect pixel correction.
Note
A mean value is calculated over the entire image that was
grabbed previously.
Definition: A defect pixel is every pixel value of this previously
grabbed image that is:
• greater than (mean value + threshold)
or
• less than (mean value - threshold)