Manual

DobotVisionStudio User Guide
Issue V4.1.2 (2022-06-08) User Guide Copyright © Yuejiang Technology Co., Ltd.
127
• Fit Mode: It has three types, including least squares, huber and tukey. The three fitting
methods only have some differences in the calculation of weight. With the increase of the number
of outliers and the distance of outliers, the least squares, Huber and Tukey can be used step by step.
Select Circle for Type
• Edge Type:
- Strongest Edge: only the edge point set with the largest gradient in the scanning range
is detected and fitted into a circle.
- Last Edge: only the edge point set with the largest distance from the center of the
circle within the scanning range is detected and fitted into a circle.
- First Edge: only the set of edge points with the smallest distance from the center of
the circle within the scanning range is detected and fitted into a circle.
• Edge Polarity:
- Dark to light: it means that the transition from the area with a low gray value to the
edge of the area with a high gray value.
- Light to dark: it means the transition from the area with a high gray value to the edge
of the area with a low gray value.
- Any means both edges are detected.
• Edge Threshold: When the select type is line, the default value is 5. The edge threshold is
the gradient threshold, ranging from 0 to 255. Only edge points whose edge gradient threshold is
greater than this value are detected. The larger the value is, the stronger the anti noise ability is, the
less the number of edges is, and even the target edge points are screened out.
• Filter Size: it is used to enhance the edge and suppress noise, and its minimum value is 1.
When the edge is blurred or there is noise interference, you can increase its value to make the
detection result more stable. If the distance between the edge and the edge is smaller than the filter
size, it will affect the accuracy of the edge location or even lose edge. This value needs to be set
based on the actual situation.
• Reject Number: It means that the number of minimum points that have high error to be
excluded from fitting. In general, if there is a great number of points excluded from fitting, its value
should be set higher. For better results, it is recommended to use it in combination with the parameter
of distance to remove.
• Initial Locating: If it is enabled, combined with the circular location sensitivity and
subsampling coefficient settings, this parameter can roughly determine that the center of the area
closer to the circle in the ROI area used as the initial circle center, which is convenient for subsequent
fine circle search. If the initial location is disabled, the ROI center is the initial circle center by
default. Under normal circumstances, the previous module of finding circle is position correction,
it is recommended to disable this parameter.
• Subsampling Coefficient: it is also called downsampling, which means that the number of
sampling points is decreased. For an image of N*M, if the subsampling coefficient is K, then one
dot is taken every K points of each row and column in the original image to create an image.
Therefore, the larger the subsampling coefficient, the more sparse the contour points will be, and
the less detailed contour will be. It is recommended that the value should not be set too large.