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

Underwater image color correction method based on GAN network

An underwater image and color correction technology, which is applied in image enhancement, image analysis, image data processing, etc., can solve problems such as blurring, not taking into account, and unable to reflect the image, so as to avoid complexity, high definition and sharpness Degree, the effect of improving image processing speed

Inactive Publication Date: 2019-08-30
XIDIAN UNIV
View PDF4 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method does not consider the influence of depth on the underwater image imaging system when making the data set, and the underwater images taken in the artificial simulation environment cannot reflect the blur caused by the depth of the camera.

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 color correction method based on GAN network
  • Underwater image color correction method based on GAN network
  • Underwater image color correction method based on GAN network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] The underwater image color correction method based on the GAN network contains the following steps (such as figure 1 shown):

[0044] Step 1. Train the network and establish a dehydration model;

[0045] Step 1.1, prepare 6500 underwater images and 6000 air images to form an unpaired data set;

[0046] Step 1.2, construct Cycle GAN network (structure such as figure 2 shown), use the unpaired dataset obtained in step 1.1 to train the Cycle GAN network to complete the style conversion from the air image to the underwater image;

[0047] Step 1.3, use the Cycle GAN network trained in step 1.2 to generate the corresponding underwater image from the image in the air in step 1.1, and the image in the air and its corresponding underwater image form a paired data set of 6000 pairs;

[0048] The generated underwater image can be used as the input of the dewatering network, and the image in the air can be used as GroundTruth (correctly labeled real image);

[0049] Step 1.4,...

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 underwater image color correction method based on a GAN network. The method comprises the steps of training a Cycle GAN network to complete the style conversion from an image in the air to an underwater image; generating the corresponding underwater images from the images in the air by using the trained Cycle GAN network, wherein the images in the air and the corresponding underwater images form a paired data set; constructing a UGAN network; using the paired data set and the actually shot underwater image as a verification set to train the UGAN network; carrying out color correction on the underwater image by using the water removal model, fusing the UV channel of the input underwater image and the Y channel of the output image in the air, and converting the fused UV channel of the input underwater image and the Y channel of the output image in the air into an RGB image to be output. According to the present invention, the data set, matched with the underwater image corresponding to the image in the air, of the generated image in the air is real and available, the image quality can be improved after the underwater image is dewatered, the processing speed is fast, and the efficiency is high.

Description

[0001] (1) Technical field: [0002] The invention relates to an image color correction method, in particular to an underwater image color correction method based on a GAN network. [0003] (two), background technology: [0004] Due to the complexity of the deep-sea shooting environment, images captured by imaging equipment often have image quality problems such as color shift, blur, and low contrast, which greatly affects the visual experience of deep-sea images and reduces the ability of imaging equipment to identify marine organisms and underwater. Effectiveness in computer vision applications such as object detection and tracking. [0005] The color correction of underwater images is mainly divided into two directions: image enhancement and image reconstruction. The algorithm for image enhancement mainly includes Retinex-based multiscale image enhancement algorithm (A Multiscale Retinex for Bridging the Gap Between Color Images and the Human Observation of Scenes), which d...

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): G06T5/00G06T7/90G06T3/00
CPCG06T7/90G06T2207/20084G06T2207/20081G06T2207/20221G06T2207/10004G06T3/04G06T5/80
Inventor 何刚卢星星李云松李磊
Owner XIDIAN 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