!
Figure 22: Without the particle filter: Violation
of the contour due to dust particles in the
vicinity of the object.
"
Figure 23: Using the particle filter: The
response to dust particles in the field is
delayed by one scan. Particles can thereby be
blanked.
3.5.8.4
Average filter
Using the mean filter, a blockwise (not moving) arithmetic average is calculated from
the configured number of scans and then outputted. The big advantage when using this
filter: Potential noise, i.e. minimal deviation of values, is reduced, which also lowers the
quantity of data.
15 m
17 m
Figure 24: Without the mean filter: The device
detects and processes all received signal val‐
ues.
Ø 16 m
Figure 25: Using the mean filter: The device
calculates an average value from several sig‐
nal values.
3.5.8.5
Median filter
This filter is suitable for excluding individual outliers from the calculation of an average
value.
The moving median filter sorts the measured values according to their size and selects
the middle value from this sequence. The values taken into consideration are defined
by a matrix, the size of which is configurable.
A 3 x 3 matrix, for example, takes into consideration 9 measured values consisting of
the distance values of a point and its two neighboring points, along with the distances
values of these points determined from the preceding and following scan. The fifth
largest measured value is outputted as the distance value.
Table 3: Example: Median for scan 2 and angle 3 (3 x 3 matrix: scan 1 to 3, angle 2 to 4)
Angle (distance values)
Scan
1
2
3
4
5
…
1
0
0
850
1100
1150
…
2
0
0
950
1200
1250
…
3
0
0
850
1150
1200
…
…
…
…
…
…
…
…
Median calculation: 0 (lowest value) | 0 | 0 | 850 |
850
(median) | 950 | 1,100 |
1,150 | 1,200 (highest value)
PRODUCT DESCRIPTION
3
8020494/1AZF/2021-05-10 | SICK
O P E R A T I N G I N S T R U C T I O N S | MRS1000
21
Subject to change without notice