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

A Uighur Image Recognition Method Based on Deep Learning

A picture recognition and deep learning technology, applied in the field of text recognition, can solve the problems of low detection recall rate, inaccurate location positioning, and unsatisfactory Uyghur language detection effect, and achieve the effect of accurate positioning and improved accuracy.

Active Publication Date: 2021-12-21
南京烽火天地通信科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Now the official Uyghur language in China is written in the Arabic alphabet, and Latin Uyghur written in the Latin alphabet is used as a supplement. The word formation and configuration components are relatively complicated. The existing general text detection technology is not ideal for the detection of Uyghur.
In addition, for Uyghur images with complex backgrounds, the existing Uyghur language detection and recognition products on the market generally have problems with low detection recall or inaccurate location positioning

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

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Uighur Image Recognition Method Based on Deep Learning
  • A Uighur Image Recognition Method Based on Deep Learning
  • A Uighur Image Recognition Method Based on Deep Learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0029] The present invention has designed a kind of Uighur picture recognition method based on deep learning, and the new neural network Smart Text Anchor Net (STAN) designed based on deep learning technology, such as figure 1 As shown, the STAN main network structure mainly includes three parts: Uyghur feature extraction network, detection network layer, classification and location optimization.

[0030] Among them, the feature extraction of Uyghur text is an important part of the Uyghur text detection network. The extraction of text features has certain particularities compared with ordinary target features. There are contextual semantic connections between different characters in text. Therefore, extracting text features also requires Reflects contextual information between characters. For this reason, we have develo...

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 present invention relates to a deep learning-based Uighur picture recognition method, a new neural network Smart Text Anchor Net (STAN) designed based on deep learning technology, aiming at the unsatisfactory detection effect of traditional text detection algorithms in complex background and changeable images The problem is that the abstract features of the image extracted by STAN are more ideal in complex background images; and the STAN network is better at detecting blurred images. Traditional algorithms are generally based on edges, connected domains or textures, and blurred images have a greater impact on such algorithms. ; Not only that, the STAN network uses smart text anchors, and re-optimizes the boundaries, which greatly improves the accuracy of Uyghur positioning, and the location of Uyghur is more accurate.

Description

technical field [0001] The invention relates to a Uyghur picture recognition method based on deep learning, which belongs to the technical field of character recognition. Background technique [0002] Uyghur is an important minority language in the Xinjiang Uyghur Autonomous Region. With the rapid development of Xinjiang, the detection and recognition technology of Uyghur characters has been widely used in document recognition, video conferencing, network security supervision, document retrieval, etc. field. The effect of Uyghur language recognition depends on the precise positioning of detection technology, which is the prior condition for Uyghur language character recognition. The detection technology of Uyghur characters in simple background is relatively mature and stable, but there are still many difficulties to be overcome in Uyghur language detection in complex background images. [0003] Traditional text detection techniques include the use of text color features, ...

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 Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06K9/32
CPCG06V30/40G06V20/62G06V30/10G06F18/214
Inventor 王康王俊涛汪洋
Owner 南京烽火天地通信科技有限公司