Adept AdeptSight User Guide - Manual (Page 490)

Search
l
Higher values reduce sensitivity to contrast. This reduces noise and the amount of
low-contrast edges.
l
Lower values increase sensitivity and add a greater amount of edge at the expense of
adding more noise. This may generate false detections and/or slow down the search
process.
Contrast threshold is the level of sensitivity that is applied to the detection of contours in the
input image. The contrast threshold can be either Adaptive or Fixed.
Adaptive thresholds set a sensitivity level based on image content. This provides flexibility to
variations in image lighting conditions and variations in contrast during the Search process.
Adaptive Low Sensitivity
Sets a low sensitivity, adaptive threshold for detecting contours. Adaptive Low
Sensitivity detects strongly defined contours and eliminates noise, at the risk of
losing significant contour segments.
Adaptive Normal Sensitivity
Sets a default sensitivity threshold for detecting contours.
Adaptive High Sensitivity
Detects the most low-contrast contours and noise.
Fixed Value
Sets an absolute value for the sensitivity to contrast. A typical situation for the
use of a fixed value is a setting in which there is little variance in lighting con-
ditions.
l
The numerical fixed value, from 1 - 255, corresponds to the minimum
step in grey-level values required to detect contours.
l
A lower value adds more contours at the expense of adding more noise,
which may generate false detections and/or slow down the search proc-
ess.
Search
Search parameters are constraints that restrict the Locator's search process.
Conformity Tolerance
Conformity Tolerance defines the maximum allowable local deviation of instance contours
from the expected model contours. Its value corresponds to the maximum distance by which
a matched contour can deviate from either side of its expected position in the model.
AdeptSight User's Guide, Version 3.2.x, Updated: 8/23/2012
Page 490