Low-illumination image enhancement method based on Retinex and deep learning

An image enhancement and deep learning technology, applied in image enhancement, image data processing, instruments, etc., to achieve outstanding effects, stable performance, and increased performance.

Active Publication Date: 2020-11-20
SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI
View PDF5 Cites 24 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

1. Establish an outdoor multiple-exposure low-light image dataset

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
  • Low-illumination image enhancement method based on Retinex and deep learning
  • Low-illumination image enhancement method based on Retinex and deep learning
  • Low-illumination image enhancement method based on Retinex and deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0049] The present invention will be further described in detail below in conjunction with the examples. The method steps are described with reference to the accompanying drawings.

[0050] 1. Dataset collection: Complete the low-illuminance and normal-illuminance image pairs in static and low-speed dynamic scenes by changing the exposure time and setting different compensation exposures, and organize and match them. The shooting equipment includes two Canon SLR cameras (Canon EOS-5D Mark II, Canon EOS-60D), two tripods, a gimbal, and a smart shooting device Kubai (used to remotely set the exposure of the SLR camera to prevent Mistakenly touched the tripod during shooting to make the scene shift).

[0051] Method 1: By adjusting the shutter time, first find a static scene in the outdoor scene, fix the tripod, set the camera to automatic photo mode, and then obtain the corresponding shutter time. Then, on this basis, remotely reduce the shutter time to shoot images with diffe...

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 a low-illumination image enhancement method based on Retinex and deep learning. An existing low-illumination data set is analyzed, low-illumination image RAW data and RGB dataare collected at the same time through multiple cameras and multiple shooting means, and the image diversity of the data set in a real scene is improved. An end-to-end improved convolutional self-encoding network structure is provided, is composed of an encoder and a decoder, and comprises an image decomposition module, an illumination adjustment module and an image reconstruction module. The image decomposition module encodes the image to obtain a reflection feature code and an illuminance hidden code. The illuminance adjusting module automatically adjusts the illuminance of the image, and the image reconstruction module recovers and reconstructs the image content. Joint loss function constraint network training, namely reconstruction loss, color loss and reflection loss, is designed. According to the method, the problems of noise removal, color distortion, image detail recovery and the like of the enhanced image in a low-light environment are solved, and the effectiveness and advancement are verified through experiments.

Description

technical field [0001] The present invention relates to a kind of image enhancement method, specifically a kind of low illumination image enhancement method based on Retinex theory and deep learning. Background technique [0002] In recent years, with the wide application of computer vision technology, image enhancement has become more and more important as a basic research work in the field of computer vision, and low-light image enhancement is one of the important research topics in the field of image enhancement. Low-light image enhancement refers to the recovery of degraded images produced by visual sensors in low-light environments through image processing, pattern recognition and other technologies. At this stage, based on different theoretical methods, low-light image enhancement has been studied by many researchers. Various deblurring, denoising and brightness adjustment techniques have been proposed, but image enhancement technology still faces many challenges. In ...

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): G06T5/00G06N3/04
CPCG06T5/002G06T5/007G06N3/045Y02T10/40
Inventor 田建东张箴荣庆轩唐延东
Owner SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI
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