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

Noise image classification method based on l2p norm robust least square method

A least square method and classification method technology, applied in the field of image processing, can solve problems such as complex steps, achieve the effect of improving classification accuracy, close connection, and ensuring authenticity

Pending Publication Date: 2021-08-27
NORTHWESTERN POLYTECHNICAL UNIV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This method has complicated steps and can only deal with specific noise in the image

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
  • Noise image classification method based on l2p norm robust least square method
  • Noise image classification method based on l2p norm robust least square method
  • Noise image classification method based on l2p norm robust least square method

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment

[0090] 1. The COIL100 data set contains 20 types of data, a total of 7200 pictures, and the size of each picture is 32×32. The gray value of each pixel of the picture is used as a feature and spliced ​​to obtain a 1024×1 vector. Therefore, the size of the data matrix 1024×7200. Normalize the data to obtain a normalized data matrix. Select 50% of the data points from each class as the training set, and the rest as the test set, then the size of the training set X is 1024×3200, and the size of the test set X t The size is 1024×3200. The size of the training set label matrix Y is 20×3200, the size of the transformation matrix W is 1024×20, and the size of the bias vector b is 20×1. Add different proportions of noise to the training set. Here, taking 10% noise as an example, the total number of noise points is 320, then the value of k is 320, and the value of the regular term parameter γ is 0.01. The initial W is calculated by the least square method 0 and b 0 .

[0091] 2. ...

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 noise image classification method based on an l2p norm robust least square method. The method comprises the steps of firstly carrying out the initialization of parameters, solving a target function of the robust least square method based on L2 and p norms through an alternating optimization method, obtaining a final classification model, and achieving the classification of noise images. According to the method, the noise and the outliers in the sample are automatically removed by adding the weights to the data points in the training set, manual error threshold selection is not needed, only the number of the noise points needs to be estimated in advance, the process is simple, the calculation amount is small, the noise image is removed while the classifier is trained, and the noise suppression capability and the image classification precision of the least square method are improved.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a noise image classification method. Background technique [0002] With the development of smart phones and information technology, more and more digital image data appear in our lives. However, the quality of digital images is affected by many factors, such as the introduction of noise and environmental interference in the process of data acquisition, transmission and processing, resulting in uneven image quality, which directly affects subsequent tasks such as classification and segmentation in image digital processing accuracy and efficiency. The noise in the image is generally divided into Gaussian noise, Poisson noise, speckle noise, salt and pepper noise, etc. At present, image denoising methods can be roughly divided into traditional model-driven denoising methods and denoising methods based on deep learning. Deep learning methods use deep neural net...

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): G06K9/62G06F17/16G06N20/00
CPCG06F17/16G06N20/00G06F18/2433G06F18/214
Inventor 王靖宇谢方园聂飞平李学龙
Owner NORTHWESTERN POLYTECHNICAL UNIV
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com