Optronis CoaXPress CamPerform Cyclone Series, Руководство пользователя

Страница: 31 / 35

-31-
correct a pixel, a valid X coordinate must be set in one of DefectPixelX0..X7 entry (valid values are 0 ..
image_width). X and Y coordinate are absolute and based on full image size even if a smaller ROI is used.
To save current correction values (i.e. pixel coordinates set in DefectPixelX0..X7 fields), user must call
DefectPixelUserFlashSave command to save all values in flash. These values will be used at next power cycle
and DefectPixelSource will return “User” as source.
To delete from flash any user data, user must call DefectPixelUserFlashErase command. Then all user correction
values are erased and DefectPixelSource will return Factory as source at next power cycle.
Defect pixel correction is controlled by XML entries:
•
DefectPixelCorrectionEnable
•
DefectPixelTestMode
•
DefectPixelSource
•
DefectPixelTotal
•
DefectPixelLineSelect
•
DefectPixelX0
•
DefectPixelX1
•
DefectPixelX2
•
DefectPixelX3
•
DefectPixelX4
•
DefectPixelX5
•
DefectPixelX6
•
DefectPixelX7
•
DefectPixelClearAll
•
DefectPixelUserFlashSave
•
DefectPixelUserFlashErase
4.3.4 Flat Field Correction (FFC)
FFC correction is a column-based correction applied on all pixels. FFC correction consists in an FPN and a PRNU
correction.
At power up, camera loads FFC flash data (= FPN and PRNU column coeff) to RAM. User can dynamically modify
column parameters using XML interface. Then, this new configuration can be saved in camera flash to be used
at next power up. If it is not saved, all user modifications will be lost after next power cycle.
When enabled, each pixel is corrected using its column parameters:
- FPN_Coeff
- PRNU_Coeff
- UserGD_Offset
- DMean_Offset
FFC correction is computed using this formula:
𝐹𝐹𝐶
𝐶𝑜𝑟𝑟𝑒𝑐𝑡𝑖𝑜𝑛
(𝑥, 𝑦) = (
𝑃𝑅𝑁𝑈
𝐶𝑜𝑒𝑓𝑓
(𝑥)
128
× (𝑃𝑖𝑥𝐼𝑛(𝑥, 𝑦) − 𝐹𝑃𝑁
𝐶𝑜𝑒𝑓𝑓
( 𝑥 ) )
) + 𝑈𝑠𝑒𝑟𝐺𝐷(𝑥) + 𝐷𝑀𝑒𝑎𝑛(𝑥)
𝑤𝑖𝑡ℎ (𝑥, 𝑦) = 𝑝𝑖𝑥𝑒𝑙 𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑎𝑛𝑑 𝑃𝑖𝑥𝐼𝑛 = 𝑝𝑖𝑥𝑒𝑙 𝑣𝑎𝑙𝑢𝑒 𝑏𝑒𝑓𝑜𝑟𝑒 𝐹𝐹𝐶 𝑐𝑜𝑟𝑟𝑒𝑐𝑡𝑖𝑜𝑛.
FFC is controlled by XML entries: