Construct Filter Feature dialog box
This command allows you to construct a filter set from a scan, certain constructed features, or another filter set. To do this, select (or key in) the input feature, the type of filter desired, and the parameters to be used with the filter. When you click the Create button, PC-DMIS applies the filter routine to the data in the input feature and displays a set marker along with the new ID for the set.
This command is typically used to smooth ball center data from a scan. PC-DMIS applies a lowpass Gaussian or another lowpass filter to smooth the data.
If you select inappropriate feature types, PC-DMIS displays an error message that says it cannot construct the feature.
To construct a filter set:
Access the Construct Filter Feature dialog box (Insert | Feature | Constructed | Filter).
Select an input for the filter set.
Select either the Polar or Linear option to smooth circle or linear data, respectively.
Select the type of filter from the Filter Type list.
Type in values for any of the filter parameters.
If you want to remove outliers before filtering, select the Outlier Removal check box and the Standard Deviation Multiple check box.
Click the Create button.
The Edit window command line for this option would read:
CONSTR/SET/FILTER,TOG1,TOG2,feat_1,
VAL1,VAL2,OUTLIER_REMOVAL/TOG3,VAL3
For example:
CONSTR/SET,FILTER,GAUSSIAN,POLAR,SCN1,
WIDTH=3,UPR =50,OUTLIER_REMOVAL/ON,3
TOG1 = Gaussian / Spline / Uniform / Triangle / Cylindrical
TOG2 = Polar / Linear
Feat_1 = This is the input feature for this filter.
VAL1 = Filter width.
VAL2 = Cutoff frequency in undulations per revolution (UPR).
TOG3 = This value switches between ON or OFF. It determines if outliers should be removed before filtering.
VAL3 = Standard deviation multiplier. If TOG3 is ON then all the points in the input that have more than this number of standard deviations from the least squares substitute feature (for example, circle, or line) will be removed before filtering.
The options supplied in this dialog box are discussed below.
More: