Classifier Class Reference

Load a saved classifier from a file. More...

Inherits object.

Public Member Functions
bool	is_minos_16bit_classifier (self)
	Returns True if the classifier was generated by Minos16Bit, False otherwise. More...
Union[Tuple[str, cvb.minos.SearchResult], List[cvb.minos.SearchResult]]	read (self, cvb.ImagePlane plane, cvb.Area2D start_aoi, cvb.Area2D ocr_aoi, Optional[float] density, int mode)
	Reads a string of characters with given mode or a list of characters with missing mode information. More...
None	save (self, str file_name)
	Write the classifier to a file. More...
cvb.minos.SearchResult	search (self, cvb.ImagePlane plane, int mode, Optional[cvb.Area2D] aoi, Optional[float] density)
	Search one object using this Minos classifier. More...
List[cvb.minos.SearchResult]	search_all (self, cvb.ImagePlane plane, int locality, Optional[cvb.Area2D] aoi, Optional[float] density=1.0)
	Search all objects using this Minos classifier. More...
cvb.minos.Classifier	set_global_advance_vector (self, cvb.Point2D advance)
	Create a new classifier, that is a copy of this classifier, but with a global advance vector applied to all classes in the classifier. More...
cvb.minos.Classifier	transform (self, cvb.Matrix2D transformation)
	Generate a new classifier by transforming this classifier with a 2 x 2 transformation matrix. More...

Properties
	comment = property
	str: Comment assigned to the classifier at generation time.
	contrast_trigger = property
	int: Trigger value for the contrast of features to be taken into account.
	extent = property
	cvb.Rect: Extent of the classes in the classifier relative to the anchor point.
	file_name = property
	str: Name of the file from which this classifier was loaded (empty string, if this image list was neither loaded nor saved since its construction).
	learn_parameters = property
	cvb.minos.LearnParameters: Name of the training set from which this classifier was generated.
	quality_measure = property
	int: Type of quality feedback from the classifier.
	threshold = property
	float: Threshold for search operations with normalized quality feedback.
	training_set_name = property
	str: Name of the training set from which this classifier was generated.

Detailed Description

Load a saved classifier from a file.

Instantiates a Classifier object.

Parameters

file_name : str Name of the file to be loaded.

Examples

Example from QmlMinos tutorial (OCR reading):

self._classifier = cvb.minos.Classifier(os.path.join(cvb.install_path(), "tutorial", "Minos", "Images", "OCR", "Training Set", "Numbers.clf"))

Member Function Documentation

is_minos_16bit_classifier()

bool is_minos_16bit_classifier

(

self

)

Returns True if the classifier was generated by Minos16Bit, False otherwise.

Returns

bool True if generated by Minos16Bit, False otherwise.

read()

Union[Tuple[str, cvb.minos.SearchResult], List[cvb.minos.SearchResult]] read	(		self,
		cvb.ImagePlane	plane,
		cvb.Area2D	start_aoi,
		cvb.Area2D	ocr_aoi,
		Optional[float]	density,
		int	mode
	)

Reads a string of characters with given mode or a list of characters with missing mode information.

Parameters

plane : cvb.ImagePlane Image plane to read on (1, 2).

start_aoi : cvb.Area2D Area in which to start the read operation (1, 2).

ocr_aoi : cvb.Area2D Area to be used for the intermediate read steps; this area will be applied relative to the expected position of the next character and is typically only a few pixels wide and high (1, 2).

density : Optional[float] Density at which to search ([0...1]) (1, 2).

mode : int Read mode (see cvb.minos.ReadMode). (2)

Returns

Union[Tuple[str, cvb.minos.SearchResult], List[cvb.minos.SearchResult]] The string that has been read by this function and the first or last character that has been read (depending on the value of the mode parameter), or the list of characters that has been read by this function.

Examples

Example from QmlMinos tutorial (OCR reading):

search_result = self._classifier.read(plane, read_aoi, OCR_AOI)
stream = ""
for i in range(len(search_result)):
    stream += search_result[i].name

save()

None save	(		self,
		str	file_name
	)

Write the classifier to a file.

Parameters

file_name : str Path to save to.

search()

cvb.minos.SearchResult search	(		self,
		cvb.ImagePlane	plane,
		int	mode,
		Optional[cvb.Area2D]	aoi,
		Optional[float]	density
	)

Search one object using this Minos classifier.

Parameters

plane : cvb.ImagePlane Image plane to search on.

mode : int Search mode to be used (see cvb.minos.SearchMode).

aoi : Optional[cvb.Area2D] Area to search in.

density : Optional[float] Density at which to search ([0...1]).

Returns

cvb.minos.SearchResult The found result (empty SearchResult if none was found).

search_all()

List[cvb.minos.SearchResult] search_all	(		self,
		cvb.ImagePlane	plane,
		int	locality,
		Optional[cvb.Area2D]	aoi,
		Optional[float]	density = 1.0
	)

Search all objects using this Minos classifier.

Parameters

plane : cvb.ImagePlane Image plane to search on.

locality : int Radius within which no better solution may exist.

aoi : Optional[cvb.Area2D] Area to search in.

density : Optional[float] Density at which to search ([0...1]).

Returns

List[cvb.minos.SearchResult] List of found results.

set_global_advance_vector()

cvb.minos.Classifier set_global_advance_vector	(		self,
		cvb.Point2D	advance
	)

Create a new classifier, that is a copy of this classifier, but with a global advance vector applied to all classes in the classifier.

Parameters

advance : cvb.Point2D Advance vector to apply globally.

Returns

cvb.minos.Classifier The new classifier.

transform()

cvb.minos.Classifier transform	(		self,
		cvb.Matrix2D	transformation
	)

Generate a new classifier by transforming this classifier with a 2 x 2 transformation matrix.

Parameters

transformation : cvb.Matrix2D Matrix with which to transform the geometry of this classifier.

Returns

cvb.minos.Classifier The transformed classifier.