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

Dermatoscope image recognition method based on StyleGANs and decision fusion

An image recognition and decision fusion technology, applied in the field of dermoscopy image recognition, which can solve the problems of many interference factors, small differences between classes, and large differences within classes.

Inactive Publication Date: 2020-04-21
CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF5 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problems of large intra-class differences and small inter-class differences, and various interference factors in dermoscopic images, the present invention develops a dermoscopic image recognition method based on StyleGANs and decision fusion

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
  • Dermatoscope image recognition method based on StyleGANs and decision fusion
  • Dermatoscope image recognition method based on StyleGANs and decision fusion
  • Dermatoscope image recognition method based on StyleGANs and decision fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] Embodiments of the present invention will be described below with reference to the accompanying drawings.

[0020] 1. Image preprocessing

[0021] In the data set, 21,701 images were selected as the training set, and 3,830 images were selected as the test set. Then process the image to uniform size. Here, in order to obtain an image suitable for the input of the convolutional neural network, the image size is uniformly set to 672*672. In order to get more image information, the image is still in color, no grayscale is needed.

[0022] 2. Data enhancement

[0023] Choose to use StyleGANs here. First, StyleGANs are trained on the training set, and then images are generated through StyleGANs, so that several realistic new samples can be obtained, from which clear and non-distorted images are selected to be added to the training set, and a larger and balanced training will be obtained. set, resulting in images such as image 3 shown.

[0024] 3. Feature learning

[...

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 provides a dermatoscope image recognition method based on StyleGANs and decision fusion. The method comprises: firstly, utilizing StyleGANs to carry out data enhancement, and then utilizing a plurality of intervention training convolutional neural networks to carry out feature learning on an image; and then putting a plurality of convolutional neural networks into a plurality of blocks, calculating the softmax average value array of the convolutional neural networks in each block, and selecting the maximum value in the array for voting; and finally, summarizing the voting resultsof the plurality of blocks, and selecting the maximum value of the voting results as a classification prediction result. According to the method, the factors of large intra-class difference, small inter-class difference, insufficient data sets and uneven distribution of the data sets under the condition of various dermatoscope images are fully considered, and the dermatoscope images can be accurately recognized.

Description

technical field [0001] The invention relates to a dermoscopic image recognition method, in particular to a dermoscopic image recognition method based on StyleGANs and decision fusion. Background technique [0002] The skin is the largest organ of the human body, and also the first line of defense of the human body, with functions such as protection, secretion, excretion and regulation of body temperature. With changes in lifestyle and environment, various skin diseases affect the normal life of human beings. Skin diseases are one of the most common diseases in humans, and melanoma skin cancer is one of the fastest growing and deadliest cancers in the world. Early diagnosis is very important for its treatment. If skin cancer is detected early and treated in time, the curative effect and prognosis are better. [0003] In order to assist doctors in diagnosis, dermoscopy is introduced clinically to improve the diagnosis of skin cell diseases, but the complexity of dermoscopy i...

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): G06K9/62G06N3/04
CPCG06N3/045G06F18/214G06F18/25G06F18/259
Inventor 龚安姚鑫杰唐永红李华昱
Owner CHINA UNIV OF PETROLEUM (EAST CHINA)
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