Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Text information feature extraction and recognition method based on Gabor filter

A feature extraction, text information technology, applied in character and pattern recognition, instruments, computer parts and other directions, can solve the problem of low efficiency of text information extraction technology, and achieve the effect of improving the recognition rate

Inactive Publication Date: 2017-05-31
HARBIN UNIV OF SCI & TECH
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to solve the problem that the extraction technology efficiency of text information in the existing video and image is low, and a kind of text information feature extraction and recognition method based on Gabor filter is proposed

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
  • Text information feature extraction and recognition method based on Gabor filter
  • Text information feature extraction and recognition method based on Gabor filter
  • Text information feature extraction and recognition method based on Gabor filter

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0018] The text information feature extraction and recognition method based on Gabor filter of the present embodiment, described text information feature extraction and recognition method are realized through the following steps:

[0019] Step 1, design Gabor filter;

[0020] Step 2, design and train the DBN classification network;

[0021] Step 3. Use the morphological method to denoise the positioned image, fill the empty area, and eliminate the isolated points to make the positioned text image more accurate, and map the final denoised text positioning binary image To the original video frame image, get the accurate text positioning area;

[0022] Step 4, perform text enhancement, binarization processing, normalization and feature extraction operations on the positioned and processed accurate text positioning area;

[0023] Step five, using the OCR recognition technology to recognize the text processed in step four.

specific Embodiment approach 2

[0025] Different from the specific embodiment one, the text information feature extraction and recognition method based on the Gabor filter of the present embodiment, the process of designing the Gabor filter described in step one refers to selecting appropriate parameters to the video frame image from 0 °, The special texture features of the characters are processed in the four directions of 45 °, 90 °, and 135 °, and four texture feature images in these four directions are obtained, the background area is suppressed, and the text texture features in the four directions are maintained. Specifically, :

[0026] The Gabor filter is regarded as a sine plane wave in the spatial domain. This sine plane wave is modulated by the Gaussian function to form a Gabor filter. Among them, the Gabor filter is determined by 7 parameters, which are the center point ,angle , mean square error and as well as and , and simplify the function of the Gabor filter by the following assumpt...

specific Embodiment approach 3

[0034] Different from the specific embodiment one or two, the text information feature extraction and recognition method based on the Gabor filter of the present embodiment, the design and training DBN classification network described in step 2 refers to using the RBM network structure to construct the DBN classification network According to the RBM network with different layers, achieve different depth DBN classification networks, compare the network structure, complexity and positioning effect of different depths, select the appropriate depth DBN classification network to process the video frame image, and locate the text area, specifically for:

[0035] The DBN network is composed of a series of probability models of restricted Boltzmann machines. The description process is as follows: Suppose there is a system S ,it has n layer , let the input be , the output is , the general process of learning is expressed as: , if the output of the system equal to its input ...

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 text information feature extraction and recognition method based on a Gabor filter to solve the problem that an existing technology for extracting text information in videos or images is low in efficiency. The method includes the following steps that firstly, the Gabor filter is designed; secondly, a DBN classification network is designed and trained; thirdly, denoising, void region filling and isolated point removing are conducted in a localized image with the morphological method to make the localized text image more precise, the denoised text localization binary image is mapped to an original video frame image, and an accurate text localization region is obtained; fourthly, text enhancement, binarization processing, normalization and feature extraction are conducted on the accurate text localization region obtained after localization and processing; fifthly, the text processed in the fourth step is recognized with the OCR technology. By means of the method, text information in videos or images can be extracted more precisely.

Description

[0001] Technical field: [0002] The invention relates to a Gabor filter-based character information feature extraction and recognition method. [0003] Background technique: [0004] In recent years, with the improvement of people's living standards and the continuous development of multimedia information technology, images and videos have become an indispensable and important information medium in people's daily life, and have also become a way of information dissemination on the Internet. In real life, multimedia content such as news, movies and TV series, and selfie videos are produced in large quantities every day. Facing such a large volume of video and images on the Internet, how to manage and use video data and retrieve important video content becomes extremely important. [0005] The text of the video has high-level semantic information. Generally, it is an aid and explanation for the video content, which is convenient for people to understand the video content and se...

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): G06K9/20G06K9/32G06K9/46
CPCG06V20/635G06V10/22G06V10/443
Inventor 刘明珠李文静郑云非
Owner HARBIN UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com