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

Multi-scale feature skin lesion deep learning recognition system based on expansion and convolution

A deep learning and recognition system technology, applied in neural learning methods, character and pattern recognition, image data processing, etc., can solve the problems of small training samples and high cost

Inactive Publication Date: 2018-04-24
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF4 Cites 32 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Moreover, the training samples that medical images can provide are small, and the samples need to be labeled by professionals, which costs a lot

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
  • Multi-scale feature skin lesion deep learning recognition system based on expansion and convolution
  • Multi-scale feature skin lesion deep learning recognition system based on expansion and convolution
  • Multi-scale feature skin lesion deep learning recognition system based on expansion and convolution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] The present invention will be described in detail below in conjunction with various embodiments shown in the accompanying drawings

[0046] The invention discloses a multi-scale deep learning skin lesion segmentation system based on dilated convolution. The specific implementation steps include:

[0047] (1) The samples in the sample library are divided into training samples and verification samples, and the pictures in the sample library are preprocessed to obtain the processed pictures.

[0048] (2) Construct a deep neural network, input the images processed in (1) into the network in batches, use the gradient descent method with momentum to optimize the network, and obtain a trained network.

[0049] (3) Perform the preprocessing as in (1) on the test sample picture to obtain the processed test picture.

[0050] (4) Input the processed test picture obtained in (3) into the trained network to obtain the predicted picture.

[0051] (5) Perform post-processing on 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 aims to solve problems of poor effect, small quantity of training samples, large difference among samples of a traditional extracted feature resorting method due to large difficulty in melanoma skin lesion segmentation and provides a multi-scale deep learning recognition system based on expansion and convolution. The system includes performing data enhancement and normalization processing on training samples and training an extracted expansion and convolution based multi-scale feature learning neural network, performing multi-threshold segmentation based on an obtained predictionprobability map, and thus implementing segmentation of a melanoma skin lesion image. Finally, the segmentation accuracy is improved.

Description

technical field [0001] The present invention relates to the field of image processing and deep learning, in particular to a skin lesion deep learning recognition system based on multi-scale features of dilated convolution. Background technique [0002] Image processing has become an attractive and promising new subject, and it is developing to a higher and deeper level. A large number of domestic and foreign literature reports have proved that people have begun to study how to use computers to interpret images and realize the use of computer vision systems to understand the external world. Many countries are vigorously exploring and researching image understanding and machine vision, and have achieved many important results. Among them, the most widely used technology in image segmentation, the most efficient and the best technology is to use deep learning to process images. [0003] With the rapid development of deep learning, most of the achievement records in the field ...

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): G06K9/62G06N3/04G06N3/08G06T7/00G06T7/136G06T7/187G06T7/194
CPCG06N3/084G06T7/0012G06T7/136G06T7/187G06T7/194G06T2207/20224G06T2207/30088G06T2207/30096G06N3/045G06F18/214
Inventor 漆进胡顺达史鹏
Owner UNIV OF ELECTRONICS SCI & TECH OF 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