Falcon4 86M Cameras
Contents
37
When off, the dark rows from the image sensor are output in the first 32 rows of the image, the
image is shifted by 32 rows, and the top 32 rows of the image are not output. This mode is used to
measure and determine if any of the black rows are defective. Defective rows can be excluded
from the dark row subtract average using the Dark Row Defect Mask.
When disabled, the normal image is passed through this module without change.
When enabled, the average of the non-defective dark rows from the current and previous frame are
averaged on a per column basis and this average is subtracted from the raw sensor data. This
results in the average output of the column to be near zero and, as a result, the Dark Row Subtract
Digital Offset feature is used to add an offset back into the data so that no zero value clipping
occurs and FPN coefficients are correctly calculated. The camera ships with a value of 50 DN and
correction coefficients are calculated with the function enabled. Users need to ensure that the FPN /
PRNU coefficients in use were calculated with the current setting of the Dark Row Subtract
Algorithm.
Flat Field Correction Algorithm Description
Flat Field Correction Algorithm (feature:
flatfieldCorrectionAlgorithm
) applies the following FFC
formula for correcting pixel values:
newPixelValue
x
,y
= (
sensorPixelValue
x
,y
–
FFCOffsetBase
x
,y – normalized
FFCOffsetDelta
x
,y
) *
FFCGain
x
,y
Where:
x
&
y
are the Flat Field Correction Pixel coordinates. (See the
flatfieldCorrectionPixelXCoordinate
and
flatfieldCorrectionPixelYCoordinate
features.)
newPixelValue
is the pixel value after Flat Field Correction is applied.
sensorPixelValue
is the pixel value before Flat Field correction is applied.
FFCOffsetBase
is one offset coefficient value to subtract from the sensorPixelValue, this
value is measured at minimal exposure time.
FFCOffsetDelta
is another offset coefficient value to subtract from the sensorPixelValue.
This value is measured at current exposure time, and is the deviation from FFCOffsetBase.
The normalization operation scales the stored
FFCOffsetDelta
by multiplying (current
integration time) / (calibration integration time). FFCOffsetDelta is measured immediately
after
FFCOffsetBase.
FFCGain
is the gain coefficient value that is multiplied with the sensorPixelValue.
The implementation of this formula requires that both the FPN and PRNU coefficient are stored in
32 bits. Internally in Falcon4, we reserve 9 bits for the FFCOffsetBase, 9 bits for FFCOffsetDelta
(FPN) coefficient and 14 bits for the FFCGain (PRNU) coefficient.
General Notes on FFC calibration
The camera comes calibrated with two factory sets, one for each shutter mode.
In addition to the
factory calibrations, the camera provides three user configurable FFC sets. These can be calibrated
and saved in the camera.
Another option is to perform the flat field correction in the frame grabber.
In either case, we recommend that you repeat the correction when a temperature change of
greater than 10 °C occurs.
For best results, ensure that:
