Operator Reference
get_texture_inspection_model_param (Operator)
get_texture_inspection_model_param — Query parameters of a texture inspection model.
Signature
get_texture_inspection_model_param( : : TextureInspectionModel, GenParamName : GenParamValue)
Description
The operator get_texture_inspection_model_param is used to query the
values of the different parameters of a texture inspection model
(TextureInspectionModel). The names of the desired parameters are
passed in the parameter GenParamName, the corresponding values are
returned in GenParamValue. It is possible to query multiple
parameters with a single call to
get_texture_inspection_model_param. All parameters which can be
manipulated by set_texture_inspection_model_param can be queried.
Please refer to the documentation of
set_texture_inspection_model_param for an explanation of the
individual parameters. It is furthermore possible to query the following
parameters:
- 'gmm_centers':
Returns the number of Gaussian distributions which were determined as optimal within the training process. The value can be used as a reference for future texture inspection models. The range of the possible number of distributions is determined by the settings of the parameters 'gmm_pmincenters' and 'gmm_pmaxcenters'. Each image pyramid level can have a different number of Gaussian distributions. By adding an index, the number of Gaussian distributions for specific pyramid levels can be queried. For example, 'gmm_centers_3' returns the number of Gaussian distributions for the third pyramid level.
- 'image_indices':
Returns the indices of the images that have been added to the texture inspection model using
add_texture_inspection_model_image. Only indices of images are returned that are currently in the texture inspection model and have not been deleted withremove_texture_inspection_model_image.- 'trained_covartype':
Returns the type of covariance matrix that was used during training. The user has the possibility to set the preferred type of covariance matrix by setting 'gmm_covartype' with
set_texture_inspection_model_param. However, if the training with the preferred covariance type fails, training is re-initialized with a less strict 'gmm_covartype'. Hence, if training with 'diag' fails, a further training with 'spherical' is attempted.
In order to avoid confusions regarding the returned tuple length,
multivalued parameters, such as 'image_indices',
'gmm_centers', 'levels' and 'novelty_threshold'
can only be queried alone i.e. without passing any other parameter to
get_texture_inspection_model_param.
For an explanation of the concept of the texture inspection see the introduction of chapter Inspection / Texture Inspection.
Execution Information
- Multithreading type: reentrant (runs in parallel with non-exclusive operators).
- Multithreading scope: global (may be called from any thread).
- Processed without parallelization.
Parameters
TextureInspectionModel (input_control) texture_inspection_model → (handle)
Handle of the texture inspection model.
GenParamName (input_control) string(-array) → (string)
Name of the queried model parameter.
Default: 'novelty_threshold'
List of values: 'gen_result_handle', 'gmm_centers', 'gmm_covartype', 'gmm_em_max_iter', 'gmm_em_regularize', 'gmm_em_threshold', 'gmm_ncomp', 'gmm_pmaxcenters', 'gmm_pmincenters', 'gmm_preprocessing', 'gmm_randseed', 'gmm_sigma', 'image_indices', 'levels', 'novelty_threshold', 'num_levels', 'patch_normalization', 'patch_rotational_robustness', 'patch_size', 'sensitivity', 'trained_covartype'
GenParamValue (output_control) string(-array) → (integer / real / string)
Value of the queried model parameter.
Example (HDevelop)
* Create texture inspection model
create_texture_inspection_model ('basic', TextureInspectionModel)
* Make this short example fast:
set_texture_inspection_model_param (TextureInspectionModel, \
'gmm_em_max_iter', 1)
* Get parameters
get_texture_inspection_model_param (TextureInspectionModel, 'patch_size', \
PatchSize)
* Read and add training images
read_image (TrainImage, 'carpet/carpet_01')
add_texture_inspection_model_image (TrainImage, TextureInspectionModel, \
Indices)
* Train the model
train_texture_inspection_model (TextureInspectionModel)
* Read and apply a test image
read_image (TestImage, 'carpet/carpet_02')
apply_texture_inspection_model (TestImage, DefectCandidates, \
TextureInspectionModel, \
TextureInspectionResultID)
Result
The operator get_texture_inspection_model_param returns the value
2 (
H_MSG_TRUE)
if the given parameters are valid and within acceptable range.
Otherwise, an exception will be raised.
Possible Predecessors
create_texture_inspection_model,
train_texture_inspection_model
Possible Successors
add_texture_inspection_model_image,
train_texture_inspection_model
See also
set_texture_inspection_model_param
Module
Matching