Method for learning text recognition, method for recognizing text using the same, and apparatus for learning text recognition, apparatus for recognizing text using the same

Abstract

A method for learning parameters used to recognize characters included in a text in a scene text image of training set is provided. The method includes steps of: (a) a training apparatus generating each feature vector corresponding to each of the segmented character images; (b) the training apparatus processing feature vectors ci+j of neighboring character images to thereby generate a support vector to be used for a recognition of a specific character image; (c) the training apparatus obtaining a merged vector by executing a computation with the support vector and a feature vector ci of the specific character image; and (d) the training apparatus (i) performing a classification of the specific character image as a letter included in a predetermined set of letters by referring to the merged vector; and (ii) adjusting the parameters by referring to a result of the classification.

Description

FIELD OF THE INVENTION[0001]The present invention relates to a method, and an apparatus for learning one or more parameters used to recognize one or more characters included in a text in a scene text image of training set, and more particularly, to the method and the apparatus for performing processes of (1) generating or allowing other device to generate each feature vector corresponding to each of segmented character images, if the segmented character images are acquired by dividing an image of the text into separate images of the characters; (2) processing or allowing other device to process feature vectors ci+j of neighboring character images to thereby generate a support vector to be used for recognizing a specific character image, on the condition that the specific character image and the neighboring character images are included in the segmented character images, wherein index j is not equal to 0 and −m≤j≤n; (3) obtaining or allowing other device to obtain a merged vector or ...

AI Technical Summary

Benefits of technology

[0010]It is another object of the present invention to provide a text recognition method with a high efficiency in identifying similar-shaped characters by performing operations on a feature vector of a specific character as a subject to be identified with a feature vector determined by referring to feature information of at least one or more of neighboring characters adjacent to the specific character.

Problems solved by technology

However, the conventional word-level processing method such as FIG. 1A may be vulnerable to variations in text length, variations in spacing between characters, and languages such as Chinese or Japanese that have no spaces in its text.

And the conventional character-level processing method such as FIG. 1B may suffer from ambiguity between similar-shaped characters, e.g., {I,l,1},{0,O},{5,S}.

As such, all the conventional text recognition approaches have such drawbacks as mentioned above.

Images