Evision
v0.1.38

View Source Evision.Text.OCRBeamSearchDecoder (Evision v0.1.38)

Summary

Types

t()

Type that represents an Text.OCRBeamSearchDecoder struct.

Functions

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table)

Creates an instance of the OCRBeamSearchDecoder class. Initializes HMMDecoder.

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table, opts)

Creates an instance of the OCRBeamSearchDecoder class. Initializes HMMDecoder.

run(self, image, min_confidence)

Recognize text using Beam Search.

run(self, image, min_confidence, opts)

Variant 1:

run

run(self, image, mask, min_confidence, opts)

run

Types

Link to this type

t()

View Source 
@type t() :: %Evision.Text.OCRBeamSearchDecoder{ref: reference()}

Type that represents an Text.OCRBeamSearchDecoder struct.

  • ref. reference()

    The underlying erlang resource variable.

Functions

Link to this function

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table)

View Source 
@spec create(
  Evision.Text.OCRBeamSearchDecoder.ClassifierCallback.t(),
  binary(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in()
) :: t() | {:error, String.t()}

Creates an instance of the OCRBeamSearchDecoder class. Initializes HMMDecoder.

Positional Arguments

  • classifier: Evision.Text.OCRBeamSearchDecoder.ClassifierCallback.t().

    The character classifier with built in feature extractor.

  • vocabulary: string.

    The language vocabulary (chars when ASCII English text). vocabulary.size() must be equal to the number of classes of the classifier.

  • transition_probabilities_table: Evision.Mat.t().

    Table with transition probabilities between character pairs. cols == rows == vocabulary.size().

  • emission_probabilities_table: Evision.Mat.t().

    Table with observation emission probabilities. cols == rows == vocabulary.size().

Keyword Arguments
Return
  • retval: OCRBeamSearchDecoder

Python prototype (for reference only):

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table[, mode[, beam_size]]) -> retval
Link to this function

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table, opts)

View Source 
@spec create(
  Evision.Text.OCRBeamSearchDecoder.ClassifierCallback.t(),
  binary(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  [{atom(), term()}, ...] | nil
) :: t() | {:error, String.t()}

Creates an instance of the OCRBeamSearchDecoder class. Initializes HMMDecoder.

Positional Arguments

  • classifier: Evision.Text.OCRBeamSearchDecoder.ClassifierCallback.t().

    The character classifier with built in feature extractor.

  • vocabulary: string.

    The language vocabulary (chars when ASCII English text). vocabulary.size() must be equal to the number of classes of the classifier.

  • transition_probabilities_table: Evision.Mat.t().

    Table with transition probabilities between character pairs. cols == rows == vocabulary.size().

  • emission_probabilities_table: Evision.Mat.t().

    Table with observation emission probabilities. cols == rows == vocabulary.size().

Keyword Arguments
Return
  • retval: OCRBeamSearchDecoder

Python prototype (for reference only):

create(classifier, vocabulary, transition_probabilities_table, emission_probabilities_table[, mode[, beam_size]]) -> retval
Link to this function

run(self, image, min_confidence)

View Source 
@spec run(t(), Evision.Mat.maybe_mat_in(), integer()) ::
  binary() | {:error, String.t()}

Recognize text using Beam Search.

Positional Arguments

  • self: Evision.Text.OCRBeamSearchDecoder.t()

  • image: Evision.Mat.t().

    Input binary image CV_8UC1 with a single text line (or word).

  • min_confidence: int

Keyword Arguments

  • component_level: int.

    Only OCR_LEVEL_WORD is supported.

Return

Takes image on input and returns recognized text in the output_text parameter. Optionally provides also the Rects for individual text elements found (e.g. words), and the list of those text elements with their confidence values.

Python prototype (for reference only):

run(image, min_confidence[, component_level]) -> retval
Link to this function

run(self, image, min_confidence, opts)

View Source 
@spec run(t(), Evision.Mat.maybe_mat_in(), integer(), [{atom(), term()}, ...] | nil) ::
  binary() | {:error, String.t()}
@spec run(t(), Evision.Mat.maybe_mat_in(), Evision.Mat.maybe_mat_in(), integer()) ::
  binary() | {:error, String.t()}

Variant 1:

run

Positional Arguments
  • self: Evision.Text.OCRBeamSearchDecoder.t()
  • image: Evision.Mat.t()
  • mask: Evision.Mat.t()
  • min_confidence: int
Keyword Arguments
  • component_level: int.
Return

Python prototype (for reference only):

run(image, mask, min_confidence[, component_level]) -> retval

Variant 2:

Recognize text using Beam Search.

Positional Arguments

  • self: Evision.Text.OCRBeamSearchDecoder.t()

  • image: Evision.Mat.t().

    Input binary image CV_8UC1 with a single text line (or word).

  • min_confidence: int

Keyword Arguments

  • component_level: int.

    Only OCR_LEVEL_WORD is supported.

Return

Takes image on input and returns recognized text in the output_text parameter. Optionally provides also the Rects for individual text elements found (e.g. words), and the list of those text elements with their confidence values.

Python prototype (for reference only):

run(image, min_confidence[, component_level]) -> retval
Link to this function

run(self, image, mask, min_confidence, opts)

View Source 
@spec run(
  t(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  integer(),
  [{atom(), term()}, ...] | nil
) :: binary() | {:error, String.t()}

run

Positional Arguments
  • self: Evision.Text.OCRBeamSearchDecoder.t()
  • image: Evision.Mat.t()
  • mask: Evision.Mat.t()
  • min_confidence: int
Keyword Arguments
  • component_level: int.
Return

Python prototype (for reference only):

run(image, mask, min_confidence[, component_level]) -> retval