Method for recognizing natural scene text in any shape

A text recognition, natural scene technology, applied in the field of natural scene text recognition, can solve the problems of inability to deal with natural scene image environmental noise interference text recognition, recognition model instability and so on

Pending Publication Date: 2021-01-05
FOSHAN NANHAI GUANGDONG TECH UNIV CNC EQUIP COOP INNOVATION INST +1
View PDF0 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most recognition models are still unstable and cannot handle text recognition problems such

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
  • Method for recognizing natural scene text in any shape
  • Method for recognizing natural scene text in any shape
  • Method for recognizing natural scene text in any shape

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0049] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. All other embodiments obtained by ordinary persons in the art without creative efforts belong to the protection scope of the present invention.

[0050] A kind of arbitrary shape natural scene text recognition method of the present invention, the overall flowchart and the detailed recognition process of scene text image are respectively as follows figure 1 and figure 2 shown, including the following steps:

[0051] S1. Acquisition of natural scene text images, making training data sets and test data sets;

[0052] S2. Construct a text recognition network model, optimize network model parameters with marked training sample data, and obtain an optimal text recognition n...

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 provides a method for recognizing a natural scene text in any shape, and the method comprises the following steps: S1, collecting a natural scene text image, and making a training data set and a test data set; S2, constructing a text recognition network model, and optimizing network model parameters by using the labeled training sample data to obtain an optimal text recognition network model; and S3, inputting a to-be-recognized natural scene text image into the trained network model to obtain a text recognition result, and combining a spatial attention mechanism in text recognition to improve spatial position information for text recognition and improve the sensitivity of an attention mechanism module to the position information, so that the method has relatively high robustness to the appearance of a text, and text recognition of any shape can be realized. The method achieves a good effect on solving the problem of detection and recognition of text instances with extreme aspect ratios or irregular shapes, and the recognition precision of the method is not affected by nearby texts or background noise.

Description

technical field [0001] The invention relates to the technical field of text detection and recognition, in particular to a text recognition method for arbitrary-shaped natural scenes. Background technique [0002] In recent years, the problem of text detection and recognition in videos and images has received more and more attention, because text is an important content to understand the entire image, and contains rich semantic information, which is the key to image content description and scene understanding Therefore, extracting text information in images has a wide application background and great research significance. The text recognition of the current rules has achieved remarkable success, such as: license plate recognition, product recognition, Internet video content security monitoring, video image retrieval, etc. Many research methods have combined recurrent neural networks and attention mechanisms into recognition models. very good result. However, most of the re...

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
IPC IPC(8): G06K9/34G06K9/20G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V10/225G06V30/153G06N3/045G06F18/253
Inventor 巴姗姗杨淑爱黄坤山
Owner FOSHAN NANHAI GUANGDONG TECH UNIV CNC EQUIP COOP INNOVATION INST
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap