Unlock instant, AI-driven research and patent intelligence for your innovation.

Low-illumination video enhancement method based on 3D convolutional neural network

A convolutional neural network and video enhancement technology, applied in the field of computer vision, can solve problems such as complex algorithms, and achieve the effect of reducing time costs and improving effects

Active Publication Date: 2020-04-21
GUANGZHOU TUWEI NETWORK TECH
View PDF12 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Firstly, the video is extracted into a continuous picture sequence, and then the algorithm is used to enhance each picture. The general algorithm of this method is relatively complicated, and the biggest defect is that it does not use inter-frame images (especially those collected by high-speed cameras) with many The same or similar information, just simply enhance each extracted picture independently

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 video enhancement method based on 3D convolutional neural network
  • Low-illumination video enhancement method based on 3D convolutional neural network
  • Low-illumination video enhancement method based on 3D convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] In order to make the above objects, features and advantages of the present invention more comprehensible, the technical solutions of the present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be pointed out that the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all those skilled in the art can obtain without creative work. Other embodiments all belong to the protection scope of the present invention.

[0037] The invention provides a method for enhancing low-light video based on a 3D convolutional neural network, comprising:

[0038] Step 1, use multiple groups of multiple continuous low-light images and corresponding normal-light images as samples to train the 3D convolutional neural network model. The input of the obtained 3D convolutional neural network model is multiple low...

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 video enhancement method based on a 3D convolutional neural network. The method comprises the following steps: step 1, training a 3D convolutional neural network model by using a plurality of groups of continuous low-illumination images as samples and corresponding normal-illumination images to obtain a 3D convolutional neural network model with the input being the plurality of low-illumination images, and the output being enhanced normal-illumination images; 2, obtaining a to-be-enhanced low-illumination video, and performing decoding and frame extraction on the to-be-enhanced low-illumination video to obtain a continuous picture sequence; 3, selecting a plurality of continuous frames of pictures in a sliding window mode, and sending the pictures to the trained 3D convolutional neural network model to sequentially obtain a continuous enhanced picture sequence; and step 4, recoding the recovered picture sequence into a video.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a low-illuminance video enhancement method based on a 3D convolutional neural network. Background technique [0002] In most computer vision technology scenarios, it is necessary to obtain clear image features in order to further analyze and process the video. However, under low-illumination conditions (such as at night), due to the weak light signal, the quality of the collected video is relatively low, and the target signal contained in the video is not clear, which further reduces the effect of subsequent video processing. Therefore, it is of great technical significance to study how to restore low-illuminance video to normal-illuminance video. [0003] The inventor found in the research that the low-illuminance video restoration in the traditional technology is essentially a low-illuminance image restoration technology. First, the video is extracted into a continuou...

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): G06T5/00G06T7/90G06N3/04G06N3/08
CPCG06T7/90G06N3/08G06T2207/10016G06N3/045G06T5/00G06T5/90
Inventor 孟焱鞠国栋沈良恒
Owner GUANGZHOU TUWEI NETWORK TECH