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

An image denoising method based on cascaded residual neural network

A neural network and neural network model technology, applied in the field of computer vision and digital image processing, can solve the problems of image noise and resolution not robust, lack of practical application value, model inaccuracy, etc., to reduce overfitting phenomenon, avoiding gradient explosion, avoiding the effect of model imprecise

Active Publication Date: 2021-07-09
SHENZHEN INST OF FUTURE MEDIA TECH +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these methods do not make full use of the NSS image blocks of noisy images and clean images at the same time, resulting in inaccurate models; in addition, the denoising process of these methods takes a lot of time, and is not robust to image noise and resolution. characteristics, lack of practical application value

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
  • An image denoising method based on cascaded residual neural network
  • An image denoising method based on cascaded residual neural network
  • An image denoising method based on cascaded residual neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] The present invention will be further described below with reference to the accompanying drawings and in combination with preferred embodiments.

[0032] The image denoising method based on the cascaded residual neural network of the present invention introduces a convolutional layer, an activation layer and a unit skip connection unit, and obtains good features on the basis of the learning ability of the convolutional layer and the screening ability of the activation layer , directly connect the input and output through the unit jump connection unit, retain more detailed information of the input image, enhance the feature extraction of the neural network model, and increase the convergence speed of the neural network model training process; thereby greatly enhancing the learning of the neural network Ability to accurately learn the mapping from noisy images to clean images to establish an input-to-output mapping, and finally predict and estimate clean images through the...

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 an image denoising method based on a cascaded residual neural network, comprising the following steps: building a cascaded residual neural network model, the cascaded residual neural network model is formed by connecting a plurality of residual units in series , wherein each of the residual units includes a plurality of convolutional layers, activation layers and unit skip connection units after each of the convolutional layers; select a training set, and set the training of the cascaded residual neural network model Parameters; according to the cascaded residual neural network model and its training parameters, the cascaded residual neural network model is trained with the goal of minimizing the loss function to form an image denoising neural network model; the image to be processed is input to the The image denoising neural network model is described, and the image after denoising is output. The image denoising method based on the cascaded residual neural network disclosed by the invention greatly enhances the learning ability of the neural network, establishes accurate mapping from noise images to clean images, and can realize real-time denoising.

Description

technical field [0001] The invention relates to the fields of computer vision and digital image processing, in particular to an image denoising method based on a cascaded residual neural network. Background technique [0002] Image denoising is a classic and fundamental problem in computer vision and image processing. It is a necessary preprocessing process to solve many related problems. Its purpose is to restore a potential clean image x from a noisy image y. The process can be expressed as: y=x+n, where n is usually considered as Additive White Gaussian (AWG), which is a typical ill-conditioned linear inverse problem. In order to solve this problem, many early methods are solved by local filtering, such as Gaussian filtering, median filtering, bilateral filtering, etc. These local filtering methods neither filter in the global scope nor consider the relationship between natural image blocks and The connection between blocks, so the obtained denoising effect is not satisf...

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 Patents(China)
IPC IPC(8): G06T5/00
CPCG06T2207/20084G06T2207/20081G06T5/70
Inventor 张永兵孙露露王好谦王兴政李莉华戴琼海
Owner SHENZHEN INST OF FUTURE MEDIA TECH
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