Operator Reference
do_ocr_word_svm (Operator)
do_ocr_word_svm — Classify a related group of characters with an OCR classifier.
Signature
do_ocr_word_svm(Character, Image : : OCRHandle, Expression, NumAlternatives, NumCorrections : Class, Word, Score)
Description
do_ocr_word_svm works like do_ocr_multi_class_svm insofar
as it computes the best class for each of the characters given by the
regions Character and the gray values Image with the OCR
classifier OCRHandle, and returns the results in Class.
In contrast to do_ocr_multi_class_svm, do_ocr_word_svm
treats the group of characters as an entity which yields a Word
by concatenating the class names for each character region. This allows
to restrict the allowed classification results on a textual level by
specifying an Expression describing the expected word.
The Expression may restrict the word to belong to a predefined
lexicon created using create_lexicon or import_lexicon,
by specifying the name of the lexicon in angular
brackets as in '<mylexicon>'. If the Expression is of
any other form, it is interpreted as a regular expression with the same
syntax as specified for tuple_regexp_match. Note that you will
usually want to use an expression of the form
'^...$'
when using variable quantifiers like '*', to ensure that the entire word
is used in the expression. Also note that in contrast to
tuple_regexp_match, do_ocr_word_svm does not
support passing extra options in an expression tuple.
If the word derived from the best class for each character does not match
the Expression, do_ocr_word_svm attempts to correct it by
considering the NumAlternatives best classes for each character.
The alternatives used are identical to those returned by
do_ocr_single_class_svm for a single character. It does so by
testing all possible corrections for which the classification result
is changed for at most NumCorrections character regions.
Note that NumAlternatives and NumCorrections affect the
complexity of the algorithm, so that in some cases internal restrictions
are made. See the section 'Complexity' below for further information.
In case the Expression is a lexicon and the above procedure did
not yield a result, the most similar word in the lexicon is returned as long
as it requires less than NumCorrections edit operations for the
correction (see suggest_lexicon).
The resulting word is graded by a Score between 0.0 (no correction
found) and 1.0 (original word correct). The Score is lowered by
adding a penalty according to the number of corrected characters and another
(minor) penalty depending on how many better classes have
been ignored in order to match the Expression:
with num_corr being the actual number of applied corrections and
num_alt the total number of discarded alternatives.
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
Character (input_object) region(-array) → object
Characters to be recognized.
Image (input_object) singlechannelimage → object (byte / uint2)
Gray values of the characters.
OCRHandle (input_control) ocr_svm → (handle)
Handle of the OCR classifier.
Expression (input_control) string → (string)
Expression describing the allowed word structure.
NumAlternatives (input_control) integer → (integer)
Number of classes per character considered for the internal word correction.
Default: 3
Suggested values: 3, 4, 5
Value range:
1
≤
NumAlternatives
NumCorrections (input_control) integer → (integer)
Maximum number of corrected characters.
Default: 2
Suggested values: 1, 2, 3, 4, 5
Value range:
0
≤
NumCorrections
Class (output_control) string(-array) → (string)
Result of classifying the characters with the SVM.
Number of elements: Class == Character
Word (output_control) string → (string)
Word text after classification and correction.
Score (output_control) real → (real)
Measure of similarity between corrected word and uncorrected classification results.
Complexity
The complexity of checking all possible corrections is of magnitude , where a is the number of alternatives, n is the number of character regions, and c is the number of allowed corrections. However, to guard against a near-infinite loop in case of large n, c is internally clipped to 5, 3, or 1 if a*n >= 30, 60, or 90, respectively.
Result
If the parameters are valid, the operator
do_ocr_word_svm returns the value 2 (
H_MSG_TRUE)
. If
necessary an exception is raised.
Possible Predecessors
trainf_ocr_class_svm,
read_ocr_class_svm
Alternatives
See also
create_ocr_class_svm,
classify_class_svm
Module
OCR/OCV