Defective Pixel Replacement (Method 4)
Important: The algorithms used exclusively with Nano firmware versions 1.00 to 1.06 are now
identified as Method 4, as described in this addendum. Nano firmware 1.07 and later implements
algorithms now identified as Method 3 and which are described in the image processing category.
The Pixel Replacement algorithm (Method 4) is based on a predefined bad pixel map (as an XML
file), either supplied by the factory (file loaded as "Factory Map") or generated by the user (file
uploaded as "User Map 1"). The number of bad pixel entries is limited and varies dependent on the
Nano model. The following XML code sample forms the template for the user to build bad pixel
maps for any of their Nano cameras.
Note: Identifying bad pixels is left to the user's discretion, but Teledyne DALSA technical support
can provide guidance.
Example User Defective Pixel Map XML File
The following example shows the required components of the defective pixel map file. Each bad
pixel position (relative to the image origin which is the upper left corner), must be identified by the
XML statement:
number
number
<DefectivePixel OffsetX="
" OffsetY="
"/>
The pixel format (whether 8, 10, 12-bit) is handled transparently, thus requires no special
consideration by the user.
This example XML listing has four "bad" pixels identified (maximum number of entries is model
dependent). The Algorithm descriptions that follow define the rules used by the Nano firmware to
replace an identified bad pixel.
<?xml version="1.0" encoding="UTF-8" ?>
<!–-Example User Defective Pixel Map -->
<!--
maximum 512 coordinates -->
<!–-filename: NanoExampleBadPixels.xml -->
<Coordinates>
<DefectivePixel OffsetX="100" OffsetY="0"/>
<DefectivePixel OffsetX="28" OffsetY="345"/>
<DefectivePixel OffsetX="468" OffsetY="50"/>
<DefectivePixel OffsetX="800" OffsetY="600"/>
</Coordinates>
An sample editable defective pixel map replacement file will be available to download with Nano
firmware files.
Monochrome Defective Pixel Replacement Algorithm
Description
The replacement algorithm follows a few basic rules as defined below, which in general provides
satisfactory results. There is no embedded intelligence to adapt the rules to avoid replacing a bad
pixel with possibly other bad data.
•
Nano Series GigE Vision Camera
Addendums
269