set_metrology_object_param — Set parameters for the metrology objects of a metrology model.
set_metrology_object_param is used to set or change the different parameters of a metrology object (see create_metrology_model for the basic principle of 2D metrology).
The metrology model is defined by the handle MetrologyHandle. The parameter Index specifies the metrology objects for which the parameters are set. The parameters of all metrology objects are set if the parameter Index is set to 'all'. All parameters can also be set when creating a metrology object with add_metrology_object_generic, add_metrology_object_circle_measure, add_metrology_object_ellipse_measure, add_metrology_object_line_measure, or add_metrology_object_rectangle2_measure. The current configuration of the metrology model can be accessed with get_metrology_object_param. All parameters that can be set with set_metrology_object_param can be reset with reset_metrology_object_param.
In the following all generic parameters with the default values are listed. But note that for a lot of applications the default values are sufficient and no adjustment is necessary. The following values for GenParamName and GenParamValue are possible - ordered by different categories:
Creating measure regions:
The value of this parameter specifies the half length of the measure regions perpendicular to the metrology object boundary. It is equivalent to the measure tolerance. The unit of this value is pixel.
List of values: 10.0, 20.0, 30.0
Default value: 20.0
Restriction: 'measure_length1' >= 1.0
The value of this parameter specifies the half length of the measure regions tangential to the metrology object boundary. The unit of this value is pixel.
List of values: 3.0, 5.0, 10.0
Default value: 5.0
Restriction: 'measure_length2' >= 0.0
The value of this parameter specifies the desired distance between the centers of two measure regions. If the value leads to too few measure regions, the parameter has no influence and the number of measure regions will be increased to the minimum required number of measure regions (circle = 3, ellipse = 5, line = 2, rectangle = 2 per side = 8). The unit of this value is pixel.
If this value is set, the parameter 'num_measures' has no influence.
List of values: 5.0, 15.0, 20.0, 30.0
Default value: 10.0
The value of this parameter specifies the desired number of measure regions. If the value leads to too few measure regions, the parameter has no influence and the number of measure regions will be increased to the minimum required number of measure regions (circle = 3, ellipse = 5, line = 2, rectangle = 2 per side = 8).
If this value is set, the parameter 'measure_distance' has no influence.
List of values: 8, 10, 16, 20, 30, 50, 100
The parameter specifies the sigma for the Gaussian smoothing. The meaning, the use, and the default value of this parameter are described with the operator measure_pos by the parameter Sigma.
The parameter specifies the minimum edge amplitude. The meaning, the use, and the default value of this parameter are described with the operator measure_pos by the parameter Threshold.
The parameter specifies the selection of end points of the edges. The meaning, the use, and the default value of this parameter are described with the operator measure_pos by the parameter Select.
The parameter specifies the use of dark/light or light/dark edges. The meaning and the use of the values 'all', 'positive', and 'negative' for the parameter 'measure_transition' is described with the operator measure_pos by the parameter Transition. Additionally, 'measure_transition' can be set to the value 'uniform'. Then, all positive edges (dark/light edges) and all negative edges (light/dark edges) are detected by the edge detection but when fitting the geometric shapes, the edges with different edge types are used separately, i.e., for each instance of a geometric shape either only the positive edges or the negative edges are used.
The measure direction within the measure regions is from the inside to the outside of the metrology object for objects of the types circle, ellipse, or rectangle. For metrology objects of the type line measure direction within the measure regions is from the left to the right, seen from the first point of the line (see RowBegin and ColumnBegin of the operator add_metrology_object_line_measure).
List of values: 'all', 'negative', 'positive', 'uniform'
Default value: 'all'
The parameter specifies the type of interpolation to be used. The meaning, the use and the default value of this parameter is described with the operator gen_measure_rectangle2 by the parameter Interpolation.
Fitting the geometric shapes:
The parameter determines what score a potential instance must at least have to be regarded as a valid instance of the metrology object. The score is the number of detected edges that are used to compute the results divided by the maximum number of measure regions (see apply_metrology_model). If it can be expected that all edges of the metrology object are present, the parameter 'min_score' can be set to a value as high as 0.8 or even 0.9. Note that in images with a high degree of clutter or strong background texture the parameter 'min_score' should be set to a value not much lower than 0.7 since otherwise false instances of a metrology object could be found.
List of values: 0.5, 0.7, 0.9
Default value: 0.7
The parameter specifies the maximum number of successfully fitted instances of each metrology object after which the fitting will stop (see apply_metrology_model). Successfully fitted instances of the metrology objects must have a score of at least the value of 'min_score'.
List of values: 1, 2, 3, 4
Default value: 1
apply_metrology_model uses a randomized search algorithm (RANSAC) to fit the geometric shapes. An edge point is considered to be part of a fitted geometric shape, if the distance of the edge point to the geometric shape does not exceed the value of 'distance_threshold'. By default, the distance threshold is set to 4 pixel.
List of values: 0, 1.0, 2.0, 3.5, 5.0
Default value: 3.5
The RANSAC algorithm estimates the number of iterations necessary for fitting the requested geometric shape. The estimation is based on the extracted edge data and the complexity of the shape. When setting the value of the parameter 'max_num_iterations', an upper limit for the computed number of iterations is defined. The number of iterations is still estimated by the RANSAC algorithm but cannot exceed the value of 'max_num_iterations'. Setting this parameter can be helpful, if the quality of the fitting is not as important as observing time limits. However, if 'max_num_iterations' is set too low, the algorithm will return low-quality or no results.
By default, 'max_num_iterations' is set to -1, indicating that no additional upper limit is set for the number of iterations of the RANSAC algorithm.
List of values: 10, 100, 1000
Default value: -1
The parameter specifies the seed for the random number generator for the RANSAC algorithm that is used by the selection of the edges the in operator apply_metrology_model. If the value of the parameter 'rand_seed' is set to a number unequal to the value 0, the operator yields the same result on every call with the same parameters, because the internally used random number generator is initialized with the value of the parameter 'rand_seed'.
If the parameter 'rand_seed' is set to the value 0, the random number generator is initialized with the current time. In this case, the results are not reproducible.
List of values: 0, 1, 42
Default value: 42
The parameter specifies the validation of the results of measurements. If the value of the parameter 'instances_outside_measure_regions' is set to the value 'false', only resulting instances of an metrology object are valid that are inside the major axis of the measure regions of this metrology object. Instances which are not valid are not stored. If the value of the parameter 'instances_outside_measure_regions' is set to the value 'true', all instances of a metrology object are valid.
List of values: 'true', 'false'
Default value: 'false'
This operator modifies the state of the following input parameter:
Handle of the metrology model.
Index of the metrology objects.
Default value: 'all'
Suggested values: 'all', 0, 1, 2
Names of the generic parameters.
Default value: 'num_instances'
List of values: 'distance_threshold', 'instances_outside_measure_regions', 'max_num_iterations', 'measure_distance', 'measure_interpolation', 'measure_length1', 'measure_length2', 'measure_select', 'measure_sigma', 'measure_threshold', 'measure_transition', 'min_score', 'num_instances', 'num_measures', 'rand_seed'
Values of the generic parameters.
Default value: 1
Suggested values: 1, 2, 3, 4, 5, 10, 20, 'all', 'true', 'false', 'first', 'last', 'positive', 'negative', 'uniform', 'nearest_neighbor', 'bilinear', 'bicubic'
If the parameters are valid, the operator set_metrology_object_param returns the value 2 (H_MSG_TRUE). If necessary, an exception is raised.
apply_metrology_model, reset_metrology_object_param, get_metrology_object_param