A composite degraded image high-quality reconstruction method based on a generative adversarial network

A degraded image and image reconstruction technology, applied in biological neural network models, image enhancement, image data processing, etc., can solve problems such as system noise, low illumination and compression distortion

Active Publication Date: 2019-04-26
BEIJING UNIV OF TECH
View PDF5 Cites 27 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the deficiencies of the prior art, and propose a composite degraded image based on a generative adversarial network architecture for low-quality images containing haze, system noise, low

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
  • A composite degraded image high-quality reconstruction method based on a generative adversarial network
  • A composite degraded image high-quality reconstruction method based on a generative adversarial network
  • A composite degraded image high-quality reconstruction method based on a generative adversarial network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] Below in conjunction with accompanying drawing of description, the embodiment of the present invention is described in detail:

[0045] A high-quality reconstruction method for composite degraded images based on generative confrontation network, the overall flow chart is attached figure 1 shown; the algorithm is divided into offline part and online part; its flow chart is attached image 3 As shown; in the offline part, the training sample set is established according to different degradation factors; for an image of size M×N, the size is first scaled to 128×128 pixels, and then the haze degradation factor and the low illumination degradation factor are added respectively , compression degrading factor, random noise degrading factor, and optical blur degrading factor to obtain training sample images, and the original image and each training sample image respectively form a training sample pair. When training the network, the training sample pairs are randomly used for...

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 discloses a composite degraded image high-quality reconstruction method based on a generative adversarial network. The method is mainly used for low-quality images with various quality reduction problems including haze, system noise, low illumination and compression distortion and the like. According to the method, a composite degraded image high-quality reconstruction method based on a generative adversarial network is established from the perspective of composite factor degraded image reconstruction, and reconstruction of a degraded image combined by factors such as haze, low illumination, compression, system noise and optical blurring can be completed; Secondly, an asymmetric generation network is adopted, so that the parameter quantity of the model is greatly reduced, andthe model is easy to train and use; Furthermore, the end-to-end idea is adopted, so that the architecture of the reconstruction system is simplified, and preprocessing and post-processing are omitted; And finally, the generation network is completely composed of convolution layers, and a composite degraded image with any size can be input for reconstruction.

Description

technical field [0001] The invention belongs to the field of digital image / video signal processing, in particular to a method for high-quality reconstruction of composite degraded images based on generating confrontation networks. Background technique [0002] In video surveillance, intelligent transportation, military imaging reconnaissance, missile precision imaging guidance, remote sensing survey, aerial surveying and mapping and other applications, outdoor vision systems are vulnerable to haze, system noise, low illumination, optical blur, compression and other factors. The random combination of these factors in a complicated way will lead to serious degradation of image quality, such as image detail loss, contrast drop, color distortion, compression block effect and other phenomena, and the subjective visual effect of the image will become very poor. At the same time, the degradation of image quality can seriously affect the effectiveness of outdoor vision systems. Wha...

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): G06K9/42G06K9/46G06N3/04G06N3/08G06T5/00G06T5/50
CPCG06N3/08G06T5/002G06T5/50G06V10/32G06V10/44G06N3/045
Inventor 李嘉锋王珂卓力张菁马春杰贾童谣
Owner BEIJING UNIV OF 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