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

Artistic text image generation method based on neural style migration

A text image and style technology, applied in the field of artistic text image generation based on neural style transfer, which can solve problems such as text differences

Active Publication Date: 2020-08-18
WUHAN UNIV OF TECH
View PDF8 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, for the style transfer of text, the method of directly using the neural network will make the shape and color of the text quite different from the actual style image

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
  • Artistic text image generation method based on neural style migration
  • Artistic text image generation method based on neural style migration
  • Artistic text image generation method based on neural style migration

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0075] The present invention will be described in detail below in conjunction with accompanying drawings and examples. The specific steps of a method for generating artistic text images based on neural style transfer in this embodiment include:

[0076] Step 1. Construct a graph-text style matching model based on the Siamese network, and obtain the style graph with the highest matching degree with the background image.

[0077] Since the size of the background image (provided by the user) and the style image (visually similar to the background image, obtained from the gallery) are different, it is necessary to consider how to extract the patch of the image while retaining as many image features as possible. . In the center of the rectangular picture, and the midpoint from the center to the diagonal, a total of five points to select a patch of 64×64 size, the background picture and the style picture take the same operation, the selection of the patch is as follows figure 1 sho...

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 an artistic text image generation method based on neural style migration, wherein the method comprises the steps of (1) image-text style matching based on a twin network; and through a matching degree recommendation algorithm based on a twin network, selecting a style graph most suitable for the background graph by utilizing a style matching algorithm; and (2) character style migration based on shape main color matching. According to the method, the structure and texture of the image are divided into two stages, two groups of generators and discriminators are guided to update parameters, and a text image pre-coloring module is introduced, so that the problem of poor background and foreground migration effect in style migration is solved, and the visual effect is greatly improved; and (3) character scale and orientation self-adaptive embedding: processing the migrated artistic text image by utilizing a character image segmentation algorithm based on distance transformation, and then carrying out self-adaptive matching on the character image and the background image by utilizing a position optimization algorithm, so that the generation efficiency of the artistic text image is improved.

Description

technical field [0001] The field of image style transfer of the present invention, in particular to a method for generating artistic text images based on neural style transfer Background technique [0002] Image style transfer is the task of transferring a style from one image to another to synthesize a new artistic image. It has a wide range of uses in visual design, such as: painting synthesis, photographic post-processing, artistic image production etc. As a kind of important semantic information, text is added to the image, and its style, position and other information affect the overall artistic visual effect of the image. Artificially generating artistic text images in specific styles requires a lot of time and effort. [0003] In recent years, there are many methods that use convolutional neural networks to extract the style features of images for transfer, and have achieved good results in image style transfer. However, for the style transfer of text, the method o...

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): G06T3/00G06T5/50G06N3/04G06N3/08
CPCG06T5/50G06N3/084G06N3/045G06T3/04
Inventor 朱安娜刘浩然
Owner WUHAN UNIV OF 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