Siemens SIMATIC MV500 - 3.3.5 Quality features during object detection

Image processing
3.3 Object recognition with PAT Genius
SIMATIC MV500
82 Operating Instructions, 03/2021, C79000-G8976-C494-05
3.3.5 Quality features during object detection
Fast feedback on the reliability of the object recognition
To obtain fast feedback about the reliability of the object recognition, diverse quality
characteristics are calculated with activated results formation. The values of the object
evaluated worst are always displayed. To view the quality characteristics of individual objects,
you can move the mouse over the respective model in the image display. The following
quality characteristics are available:
Table 3- 18 Quality characteristic of object recognition
Quality characteristic
Value output
Conformity (LFI)
The output value indicates the normalized, average deviation from
matched model features and the currently found features of the
object instance.
• A value of 100 means an average deviation of zero.
• Accordingly, the value 0 signifies: No match.
Match (LMA)
The output value indicates the proportion of matched model features
(edges).
• A value of 100 means that 100% of the required model features
were successfully matched.
• A value of 0 means that none of the required model features
were matched.
Clear percentage (LCL)
The output value describes the clear percentage of the instance.
• A value of 100 means that the found object instance is free of
disturbances.
Note
This value is calculated only if "Minimum clear percentage" is activat-
ed in the "Recognition" group of the relevant Locator step. Otherwise,
the value 100 is output.