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

A Method of Image Classification Based on Multi-resolution

A classification method and multi-resolution technology, applied in the field of image classification based on multi-resolution, which can solve problems such as overfitting and network degradation

Active Publication Date: 2021-03-16
CHINA JILIANG UNIV
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But as the network deepens, the problems of overfitting and network degradation become more serious

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 Method of Image Classification Based on Multi-resolution
  • A Method of Image Classification Based on Multi-resolution
  • A Method of Image Classification Based on Multi-resolution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] Such as figure 1 with figure 2 As shown, the present invention discloses a multi-resolution-based image classification method, and the specific implementation manner of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0038] Step 1: Dataset preparation. Select a data set containing 10 categories, and each category of pictures contains 3 resolutions, namely 64×64, 128×128, and 256×256. Each resolution has a similar number of pictures. Divide the images into two parts, training set and test set.

[0039] Step 2: Make image labels. Since there are 10 categories, there are 10 categories of labels, expressed as l(l∈(1,10)). In the training set, the labels of pictures of the same category are consistent, and the labels of different categories cannot be the same.

[0040] Step 3: Build a picture database. When training a deep convolutional neural network, you need to input training data, and use the prepared data s...

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 classification method based on multi-resolution, the purpose of which is to use different deep neural network structures to solve multi-resolution input problems, and use deep learning technology to find a classification label from the classification label set and classify The effect of label assignment to the input image. The technical key lies in (1) adopting the method of skipping part of the pooling layer to realize multi-resolution image classification; (2) adopting the method of specifying position input to realize multi-resolution image classification. (3) According to the different characteristics of the features obtained by different layers of the convolutional network, different utilization methods are used for the features of different layers. The present invention inputs any test picture into the trained neural network model, and the output of the neural network is the category of the picture. On the premise of not unifying the size of the input picture, the invention maintains the quality of the original picture, does not add any noise, and effectively realizes multi-resolution image classification.

Description

technical field [0001] The invention belongs to the field of image processing, in particular to an image classification method based on multi-resolution. Background technique [0002] As an important part of the computer vision field, image classification can effectively analyze the content of images, obtain key information in images, and give correct judgments. Image classification seems to be the simplest problem, but it is also a core problem. Its importance is also reflected in some other computer vision problems, such as object positioning and recognition, image content segmentation, etc., which have great impact on real work life and social development. is of great significance. [0003] Image classification focuses on global statistical information, and for image data, deep learning has excellent modeling and feature extraction capabilities, and has been widely used in theoretical analysis and practical applications of image object classification. Kong proposed to b...

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): G06K9/62G06N3/04G06F16/583
CPCG06F16/583G06N3/045G06F18/24G06F18/214
Inventor 章东平倪佩青胡葵杨力张香伟
Owner CHINA JILIANG 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