A method and system for automatic identification of inverted Chinese character verification codes

Abstract

The invention discloses a method and system for automatic recognition of inverted Chinese character verification codes. The method includes: processing the verification code image to be recognized to obtain a target verification code image, and comparing the target verification code image with the left and right borders of each character in the target verification code image. Segment the verification code image to obtain each single-character image; generate label information corresponding to the target verification code image, the label information is expressed in binary, 0 indicates that the characters in the single-character image are normal, 1 indicates that the characters in the single-character image are inverted; The length of the information is passed to the trained inverted Chinese character automatic recognition model as a parameter, and the inverted Chinese character automatic recognition model sets the output number of neurons to be consistent with the length of the label information; the sum of cross entropy is used as the trained inverted Chinese character automatic recognition model. Identify the loss function of the model, train the loss function until the loss function converges, and output the recognition result. It can realize the automatic recognition of the inverted Chinese character verification code.

Description

technical field [0001] The invention belongs to the technical field of automatic recognition, and more specifically relates to a method and system for automatic recognition of inverted Chinese character verification codes. Background technique [0002] With the rapid development of the Internet, while the Internet has brought great convenience to people's lives, its security issues have become increasingly prominent. As a widely used verification method, network verification code plays an important role in network security. [0003] At present, the recognition of character-type verification codes is mainly the recognition of text. The main process is as follows: prepare the original picture material; picture preprocessing, character cutting, picture size normalization; picture character labeling; character picture feature extraction; generating features and Mark the corresponding training data set; train the feature-labeled data to generate a recognition model; use the reco...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the purpose of the present invention is to provide a method and system for automatic recognition of inverted Chinese character verification codes, thereby solving the technical problem that traditional verification code recognition technologies cannot recognize inverted Chinese characters

Images