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

Graphic pattern text detection method based on deep learning

A text detection and deep learning technology, applied in the field of graphic pattern text detection based on deep learning, can solve the problems of complex background and time-consuming

Active Publication Date: 2015-07-22
ZHEJIANG UNIV
View PDF3 Cites 48 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The background in natural images is complex, and the positions of graphics, patterns, and texts appear randomly, and their sizes are random. Due to different shooting angles, changes such as tilt, rotation, and perspective transformation may also occur. This is the main difficulty in the detection of graphics, patterns, and texts in natural images.
In addition to the complexity of the problem itself, there are not many marked samples for the problem of graphic pattern text detection, and the marked samples must be manually marked, which is very time-consuming

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
  • Graphic pattern text detection method based on deep learning
  • Graphic pattern text detection method based on deep learning
  • Graphic pattern text detection method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0074] In this embodiment, a certain group of images with text position marks is trained, and a group of images without marks is used for text positioning. In the embodiment of the present invention, the method includes the following steps:

[0075] Step 1: Preprocessing: convert the input color image into a grayscale image, perform multi-resolution decomposition, and output images of different resolutions;

[0076] Step 2: Feature extraction: For each resolution image output in step 1, feature maps are extracted through a deep convolutional self-encoding network. Among them, the deep convolutional autoencoder network is obtained through training;

[0077] Step 3: Sparse coding solution: All the feature maps extracted in step 2 are upsampled to the size of the original image. The group of feature maps is divided into blocks, and each block is classified and identified through a sparse dictionary and a linear classifier.

[0078] Step 4: Text positioning: perform regional fus...

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 graphic pattern text detection method based on deep learning. The method includes firstly, training a depth convolution self-encoding network by combining graphic pattern text samples, than adopting marked samples, and classifying through a sparse dictionary; extracting graphic pattern texts from a sample library, rotating, shifting and transmitting, and combining the graphic pattern texts with pure background graphics; adopting a combined sample seat, establishing a depth convolution self-encoding network, and learning characteristic templates in layered training and entire optimizing manners; performing characteristic extraction on the characteristic templates acquired by deep network learning according to the acquired marked samples; sampling the extracted characteristics in the size of the original graphics, adopting single blocks as identifying units, and training the sparse dictionary and a classifier; after training, performing multi resolution decomposition on the graphics to be processed, utilizing the characteristic templates to extract characteristics, and utilizing the sparse dictionary to acquire results in a classified manner.

Description

technical field [0001] The invention relates to a graphic pattern text detection algorithm in the technical field of computer vision, in particular to a graphic pattern text detection method based on deep learning. Background technique [0002] With the development of the Internet and the continuous improvement of network bandwidth, images and videos are easy to understand and conform to the fast pace of life of modern people, gradually replacing text as the main carrier of information dissemination. Due to the limitation of text length on microblogging websites, a combination of graphics and texts is usually required to publish richer content. At the same time, the vigorous development of picture sharing websites has brought new opportunities and challenges to the field of image retrieval and understanding. [0003] Image retrieval generally has two directions, one is to directly match based on the features of the image, and the other is to mark the image first and then ret...

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): G06K9/66G06T7/00
Inventor 于慧敏李天豪
Owner ZHEJIANG UNIV
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