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

Underwater image restoration method based on fusion countermeasure network

An underwater image and network technology, applied in image enhancement, image data processing, instruments, etc., can solve the problems of poor underwater image restoration effect, complicated underwater image restoration and enhancement methods, etc., achieve bright contrast, improve image restoration, etc. Performance, Guaranteed Content and Effect of Structural Similarity

Inactive Publication Date: 2018-12-21
OCEAN UNIV OF CHINA
View PDF3 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention provides an underwater image restoration method based on a fusion confrontation network to solve the existing technical problems such as complex restoration and enhancement methods for underwater images, the need to obtain a large number of underwater condition parameters, and poor underwater image restoration effects.

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
  • Underwater image restoration method based on fusion countermeasure network
  • Underwater image restoration method based on fusion countermeasure network
  • Underwater image restoration method based on fusion countermeasure network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0034] The embodiments of the present application are preferred embodiments of the present application.

[0035] An underwater image restoration method based on a fusion confrontation network, using unpaired turbid underwater images and clear underwater images (the unpaired here refers to the number of clear underwater images corresponding to several turbid underwater images, but not It needs to satisfy the one-to-one correspondence relationship between image structure content and quantity), through the unsupervised deep learning method, the underwater image in the turbid domain is restored to the underwater image in the clear domain, and the restored clear underwater image has more natural color and better contrast. Stronger, clearer content, including the following steps:

[0036] Collect clear underwater images, construct a clear underwater image data set, and use the clear underwater image data set as a real clear domain underwater image;

[0037] Collecting turbid underw...

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 underwater image restoration method based on a fusion countermeasure network, wherein the clear underwater image data set is taken as a real clear underwater image. Taking the turbidity underwater image data set as a real turbidity underwater image; Obtaining a transmittance map of the real turbidity region underwater image through a dark channel prior algorithm, and obtaining depth information of the real turbidity region underwater image through the transmittance map; A multi-layer countermeasure neural network model is constructed, the real turbidity underwater image, its depth information and the real clear underwater image are inputted into the network model, and the real turbidity underwater image is converted into a synthesized clear underwater image through training and iterative feedback. In addition, the restored underwater image is compared with other methods, which provides a basis for the further study of underwater vision task.

Description

technical field [0001] The invention relates to the technical field of underwater computer vision, in particular to an underwater image restoration method based on a fusion confrontation network. Background technique [0002] Underwater observation is an important prerequisite for us to explore the ocean and develop and utilize underwater resources. Many underwater devices are available for underwater observations, however the quality of raw underwater images is usually not high enough to be selected for further research. Due to the absorption and scattering of underwater light and the influence of water quality, the underwater images we get generally have disadvantages such as color distortion, low contrast and noise. [0003] Currently, underwater image restoration is still a challenging subject. In order to solve the problem of underwater image restoration, some traditional methods do not depend on the physical model, and improve the visual quality of the image by adjus...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06T5/00
CPCG06T5/00
Inventor 俞智斌卢婧宇郑海永郑冰
Owner OCEAN UNIV OF CHINA
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