Underwater image enhancement method based on multi-scale dense generative adversarial network and training method of network model

A technology of network model and training method, applied in biological neural network model, image enhancement, image data processing, etc., can solve problems such as difficulty in implementation, poor generalization, and poor implementation effect, so as to ensure diversity and good Restoring, increasing clarity effects

Pending Publication Date: 2022-06-24
NANJING UNIV OF INFORMATION SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to complex underwater physical and optical factors and insufficient hardware equipment, traditional methods are difficult to implement, and the effect is not very good. At the same time, due to the lack of rich training data, these methods are widely used in different underwater environments. less chemical

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 enhancement method based on multi-scale dense generative adversarial network and training method of network model
  • Underwater image enhancement method based on multi-scale dense generative adversarial network and training method of network model
  • Underwater image enhancement method based on multi-scale dense generative adversarial network and training method of network model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] This embodiment provides a training method for a multi-scale dense generative adversarial network model (hereinafter referred to as the UWGAN-VGG model), and the trained UWGAN-VGG model can effectively enhance underwater images. The steps of UWGAN-VGG model training are as follows.

[0060] 1. Prepare the training data set

[0061] The training data set used in this embodiment is mainly generated by CycleGAN, which is essentially two mirror-symmetric GAN networks, and the cycle-generating adversarial network CycleGAN is an unsupervised network model that does not require paired data sets. The training data set includes the original image in the undistorted air and the distorted image mapped with the original image. These two sets of images do not need to have any correlation, that is, two sets of asymmetric data, and the training data set is asymmetric data. Set; the original undistorted images come from a subset of the ImageNet image set, and the distorted underwater ...

Embodiment 2

[0101] This embodiment provides an underwater image enhancement method based on a multi-scale dense generative adversarial network, and the multi-scale dense generative adversarial network adopts the UWGAN-VGG model trained in Embodiment 1. The enhancement method of the underwater image is as follows:

[0102] Get the original underwater picture;

[0103] The underwater original image is input into the trained multi-scale dense generative adversarial network model for image enhancement processing, and the underwater enhanced image is output.

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 enhancement method based on a multi-scale dense generative adversarial network and a training method of a network model, the multi-scale dense generative adversarial network model comprises a generator network and a discriminator network, and the generator network comprises a jump connection block composed of a multi-scale dense feature extraction MSDB module. In addition to unsaturated loss, L1 distance and gradient loss, the model also adds VGG perceptual loss to enhance the structure of the UWGAN, the true value of the ground can be more clearly reserved through the unsaturated loss, the L1 distance and the gradient loss, the perceptual loss is used for comparing the difference of image feature space, so that the network can better recover the detail information of the underwater image, and the accuracy of the image feature space is improved. And underwater image enhancement is carried out to increase the definition of the image.

Description

technical field [0001] The invention belongs to the technical field of underwater image processing, and in particular relates to an underwater image enhancement method based on a multi-scale dense generative confrontation network and a training method of a network model. Background technique [0002] In recent years, in order to build a marine power, exploit marine resources, and build a community with a shared future for the ocean, the proportion of applications based on information technology in my country's marine field has increased significantly, and the observation of underwater images has become the focus of the development of the marine industry. Underwater images play an important role in the fields of underwater robot intelligent fishing, underwater navigation, and underwater engineering detection. Clear underwater images can be used for large-scale underwater projects such as underwater target detection and tracking, target classification, and underwater mineral ex...

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): G06T5/00G06K9/62G06N3/04G06N3/08G06V10/44G06V10/774G06V10/80
CPCG06N3/08G06N3/048G06N3/045G06F18/214G06F18/253G06T5/00
Inventor 刘磊陈海秀颜秋叙金肃钦刘奇巩大康何珊珊陆康
Owner NANJING UNIV OF INFORMATION SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap