License OCR recognition method

Abstract

The invention provides a license OCR recognition method in the technical field of image recognition, and the method comprises the following steps: S10, obtaining a large number of license sample pictures and license information, creating a training dictionary and a plurality of license background pictures, and generating a license training picture set based on the license information, the training dictionary and the license background pictures; S20, creating a license classification model, and training the license classification model by using the license sample pictures; S30, creating a plurality of text recognition models based on the license type, and training the text recognition models by using the license training picture set; and S40, classifying and identifying the to-be-identified license by utilizing the trained license classification model and text identification model. The method has the advantages that the certificates are automatically classified and recognized, and the certificate recognition precision is greatly improved.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to a certificate OCR recognition method. Background technique [0002] OCR (Optical Character Recognition, Optical Character Recognition) technology is to convert the text of various bills, newspapers, books, manuscripts and other printed materials into image information through optical input methods such as scanning, and then use text recognition technology to convert the image information into usable The input technology of computer language. [0003] With the continuous development of image sensors, especially the exponential increase in the number of various mobile phones and professional (such as security) cameras, computer image data is increasing rapidly; but the image quality is reduced compared to traditional scanners or various professional cameras , making the traditional Chinese character OCR technology face the low quality of source image data and serious pol...

AI Technical Summary

Problems solved by technology

[0003] With the continuous development of image sensors, especially the exponential increase in the number of various mobile phones and professional (such as security) cameras, computer image data is increasing rapidly; but the image quality is reduced compared to traditional scanners or various professional cameras , making the traditional Chinese character OCR technology face the low quality of source image data and serious pollution, resulting in a sharp drop in recognition rate

[0004] Chinese character content recognition (Chinese character OCR) of computer images is a difficult problem in image recognition. Compared with English character recognition, Chinese characters are susceptible to interference and difficult to distinguish due to their large number and high similarity of basic characters.

For the license, it will also be affected by the complex shading of the license itself, which reduces the accuracy of recognition

For various bills taken by mobile phones, the traditional OCR manufacturers do not have ideal recognition results. Although the end-to-end OCR recognition method based on deep neural networks has achieved good results for Western characters, Chinese characters require a lot of training due to their large base. The data set far exceeds the western character set, so the performance of Chinese character OCR on the open AI platform is quite unsatisfactory on poor images, and the end-to-end deep neural network has natural misrecognition and is vulnerable

[0005] For the identification of licenses, currently only the model and api interface can be customized for each license recognition scenario, and the api interface needs to be designed separately for each license and even new training data needs to be produced for new training, which increases the computing resources. The waste and redundancy of memory usage also increase the complexity of the user's use of the ID card identification function

Images