Unlock instant, AI-driven research and patent intelligence for your innovation.

Water book character recognition method based on CNN structure neural network

A neural network and text recognition technology, applied in the field of text recognition, can solve problems such as low accuracy of water script text, achieve the effect of reducing over-fitting phenomenon, reducing the amount of parameters, and enhancing features

Pending Publication Date: 2019-10-18
贵州工业职业技术学院
View PDF3 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is to provide a water script recognition method based on a CNN structure neural network to solve the problems of low accuracy in identifying water script characters by conventional text recognition technology in the prior art

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] Step 1. Collect samples of water script;

[0027] In order to solve the translation text of Shuishu characters, we chose 120 pages of "Shuishu Dictionary" as the basic sample;

[0028] Step 2. Use a neural network model based on the CNN structure to extract and classify the water book text samples;

[0029] Image feature extraction refers to obtaining various metrics or attributes useful for classification from the object itself. According to the feature vector obtained by feature extraction, the object is assigned a category mark, so that the analysis sample is divided into n categories. It is generally believed that two objects are similar because they have similar features, so samples with similar features belong to the same category.

[0030] In traditional machine learning methods, most of them manually extract the features of the image to be classified, and then put the features into common classifiers (such as SVM, decision tree, random forest) for classification, and ob...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a water book character recognition method based on a CNN structure neural network. The water book character recognition method comprises the steps of 1, collecting a water bookcharacter sample; 2, carrying out feature extraction and classification on the water book character sample by adopting a neural network model based on a CNN structure; 3, positioning and detecting the characters of the water book by adopting a YOLO algorithm. The problem that in the prior art, a conventional character recognition technology is low in accuracy when used for recognizing water bookcharacters is solved.

Description

Technical field [0001] The invention belongs to character recognition technology, and particularly relates to a water book character recognition method based on a CNN structure neural network. Background technique: [0002] In the field of computer vision, object recognition and positioning has always been an important research direction. The recognition of water script characters is also a problem in this category. Because water script characters are pictographs, they are non-standardized characters. In the sample, almost all are handwritten water book materials. The same character written by different people may vary greatly. There are also many water book characters that are mixed with other characters (for example: Chinese characters). Accurately identifying Shuishu characters in the literature without misrecognizing other characters or patterns, which puts forward high requirements for character segmentation, positioning and detection, and character feature extraction and cl...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/20G06K9/34G06N3/04G06N3/08
CPCG06N3/084G06V30/333G06V30/36G06V10/22G06V30/153G06N3/045
Inventor 丁琼
Owner 贵州工业职业技术学院