Page 54
Advanced Settings
confocalDT 24x1
Processing sequence:
1. Video averaging
2. Unlinearized distances
3. Linearizing the distances
4. Refractive index correction of the distances
5. Troubleshooting when there is no valid measured value
6. Spike correction of the distances
7. Difference generation for thicknesses
8. Measured value averaging
9. Statistics
6.3.2
Video Averaging
The following video graphs can be averaged successively and pixel by pixel in the sen-
sor. In the web browser, under
Video signal
, you can see the effect of the various
settings.
Video averaging is especially recommended for very small video signal peaks to help
reduce the threshold and to achieve a greater number of valid readings.
The video averaging must be ignored in the two times alternating exposure mode.
6.3.3
Measurement Averaging
Measurement averaging is performed after measurement values have been calculated,
and before they are issued or processed through the relevant interfaces.
Measurement averaging
- improves the resolution
- allows masking individual interference points, and
- 'smoothes' the reading.
i
Linearity is not affected by averaging. Averaging has no effect on measuring rate
and output rate.
The internal average value is re-calculated for each measuring cycle.
i
The defined type of average value and the number of values must be stored in the
controller to ensure they are hold after it is switched off.
Controller IFC24x1 is delivered with “moving average, averaging value = 1” as factory set-
tings, ie. averaging is not enabled by default.
Moving average
The definable number N for successive measurements (window width) is used to calcu-
late the arithmetic average M
mov
according to the following formula:
MV (k)
k=1
N
N
M =
mov
MV = measured value
N = averaging value
k = continuous index (in the window)
M
mov
= average value or output value
Each new measured value is added, and the first (oldest) value is removed from the av-
eraging (from the window). This produces short response times for measurement jumps.
Example
: N = 4
... 0, 1, 2, 2, 1, 3
2, 2, 1, 3
4
= M (n)
mov
... 1, 2, 2, 1, 3, 4
2, 1, 3, 4
4
= M (n+1)
mov
Measured values
Output value
