Class cv::text::OCRBeamSearchDecoder::ClassifierCallback#

Callback with the character classifier is made a class. View details

#include <opencv2/text/ocr.hpp>

Collaboration diagram for cv::text::OCRBeamSearchDecoder::ClassifierCallback:

cv::text::OCRBeamSearchDecoder::ClassifierCallback Node1 cv::text::OCRBeamSearchDecoder ::ClassifierCallback   + ~ClassifierCallback() + eval() + getStepSize() + getWindowSize()

cv::text::OCRBeamSearchDecoder::ClassifierCallback Node1 cv::text::OCRBeamSearchDecoder ::ClassifierCallback   + ~ClassifierCallback() + eval() + getStepSize() + getWindowSize()

Detailed Description#

Callback with the character classifier is made a class.

This way it hides the feature extractor and the classifier itself, so developers can write their own OCR code.

The default character classifier and feature extractor can be loaded using the utility function loadOCRBeamSearchClassifierCNN with all its parameters provided in opencv/opencv_contrib.

Constructor & Destructor Documentation#

~ClassifierCallback()#

cv::text::OCRBeamSearchDecoder::ClassifierCallback::~ClassifierCallback()

Member Function Documentation#

eval()#

void cv::text::OCRBeamSearchDecoder::ClassifierCallback::eval(
InputArray image,
std::vector< std::vector< double > > & recognition_probabilities,
std::vector< int > & oversegmentation )

The character classifier must return a (ranked list of) class(es) id(‘s)

Parameters

  • image — Input image CV_8UC1 or CV_8UC3 with a single letter.

  • recognition_probabilities — For each of the N characters found the classifier returns a list with class probabilities for each class.

  • oversegmentation — The classifier returns a list of N+1 character locations’ x-coordinates, including 0 as start-sequence location.

getStepSize()#

int cv::text::OCRBeamSearchDecoder::ClassifierCallback::getStepSize()

getWindowSize()#

int cv::text::OCRBeamSearchDecoder::ClassifierCallback::getWindowSize()

Source file#

The documentation for this class was generated from the following file: