Pike Technical Manual
V5.2.0
357
Configuration of the camera
Reading or writing defect pixel correction data from/into the
camera
Accessing the defect pixel correction data inside the camera is done through the
GPDATA_BUFFER. Because the size of the GPDATA_BUFFER is smaller than the
whole defect pixel correction data the data must be written in multiple steps.
Defect pixels and columns are saved in two adjacent memory chunks, first the
defect pixel chunk followed by defect columns.
DPC-Memory alignment: {defect_pixel_1, defect_pixel_2 … defect_pixel_n,
defect_column_1, defect_column_2 … defect_column_n}.
One defect pixel is saved as two 16-bit values (X, Y), so the size of defect pixel
chunk equals
NumberDefectPixel
* 4 (
it is also the offset of defect column
data
).
One defect column is saved as three 16-bit values (X, Y, Height), so the size of
defect column chunk equals
NumberDefectColumn * 6
To write DPC coordinates:
1.
Query the limits and ranges by reading DEFECT_PIXEL_CORRECTION_INFO
and GPDATA_INFO.
(Note: If the list is empty, you don‘t have to write the DPC pixels. In this
case do the following: in DEFECT_PIXEL_CORRECTION_MEM set the
NumberDefectColumn and NumberDefectPixel to 0.)
2.
Set EnableMemWR to true (1).
3.
Update NumberDefectColumn and NumberDefectPixel in DEFECT_PIX-
EL_CORRECTION_MEM to the new values.
4.
Write n DPC data bytes to GPDATA_BUFFER (n might be lower than the size
of the GPDATA_BUFFER)
5.
Repeat step 4 until all data is transferred.
0xF1000468 DEFECT_PIXEL_CORRECTION_INFO Presence_Inq
[0]
Indicates presence of this
feature (read only)
Version
[1..3]
Feature version
---
[4..19]
reserved
MaxDPCTable
Size
[20..31] Maximum size of DPC table
(in 128 Byte Blocks)
Note
Pixel coordinates and column coordinates must be sorted
ascending, by X as primary and Y as secondary sorting-key.
Register
Name
Field
Bit
Description
Table 193: Advanced register:
Defect pixel correction
