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

Generative adversarial network-based old drawing repair method

A repair method and drawing technology, applied in the field of image processing, to achieve the effect of being easy to read manually

Pending Publication Date: 2022-03-22
NANJING INST OF MECHATRONIC TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, for damaged and blurred old drawings, we can only rely on manual repair and redrawing after reading and understanding

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
  • Generative adversarial network-based old drawing repair method
  • Generative adversarial network-based old drawing repair method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] The technical solutions of the present invention will be further elaborated below in conjunction with the description of the drawings and specific embodiments.

[0029] Such as figure 1 Shown, method of the present invention comprises the steps:

[0030] 1. Graphical model:

[0031] a) Network selection - ESRGAN.

[0032] b) Dataset - The dataset for generating an adversarial network does not need to be labeled, just provide fuzzy and corresponding clear drawings.

[0033] i. Collect legible drawings and scan them at a uniform resolution;

[0034] ii. Generally, the drawing size is much larger than the picture, so the drawing needs to be divided into multiple sub-pictures. The image segmentation size can be adjusted according to the computing power and actual effect;

[0035] iii. Blur the segmented image. One or more blurring methods can be used, such as directly interpolating and enlarging the picture (compressing the image and interpolating and enlarging it to ...

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 relates to an old drawing repairing method based on a generative adversarial network, which comprises the following steps of: collecting drawings which can be read clearly, scanning the drawings into digital images with unified resolution, and respectively constructing a graphic model and a text model based on the generative adversarial network, constructing a training set by using the positioned text region to train a generative adversarial network model, and obtaining a graphic model; finally, the old drawing is input into the graphic model and the text model, output results are combined, and the repaired old drawing is obtained. According to the method, a large batch of old drawings can be automatically and quickly processed and deblurred, so that the drawings are clear and readable, and the effect of text clearness is improved as much as possible on the premise of ensuring the clearness of the drawings. The processed drawings are convenient to read manually, and a foundation is laid for automatic conversion and key information extraction in the next step.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an old drawing restoration method based on a generative confrontation network. Background technique [0002] When transforming existing factories and facilities, as well as digitally transforming and rebuilding factories, the accuracy of existing drawings and documents is crucial. For many old factories, most of their drawings are kept in the archives on paper. After being stored for a long time, the drawings are often stained or aged, and are no longer clear, especially after scanning, it is even more difficult to read. In order to carry out the transformation and reconstruction smoothly, it often takes a lot of manpower and material resources to check, check and redraw the content on the drawings or documents. [0003] Generative Adversarial Network (GAN) is a new artificial intelligence technology developed on the basis of convolutional neural network. ...

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): G06V30/40G06V30/19G06K9/62G06T5/00G06T7/11G06T7/12G06T3/40G06N3/04G06N3/08
CPCG06T7/11G06T7/12G06T3/4007G06N3/04G06N3/08G06T2207/20081G06T2207/20084G06T2207/30176G06F18/214G06T5/73G06T5/77
Inventor 张萌
Owner NANJING INST OF MECHATRONIC 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