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

TCM complexion automatic sorting method applying superficial layer neural network

A neural network and automatic classification technology, applied in image analysis, image data processing, instruments, etc., can solve the problems of difficult training, small number of samples in high-quality complexion data sets, and inability to apply TCM facial complexion classification well, so as to achieve strong Robustness, the effect of improving classification accuracy

Active Publication Date: 2017-12-26
BEIJING UNIV OF TECH
View PDF4 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the field of face-to-face consultation in traditional Chinese medicine, on the one hand, high-quality medical data samples are seriously scarce; on the other hand, in the process of data collection, due to interference from external factors, the number of correctly labeled high-quality complexion data sets obtained is relatively small. It is difficult to meet the needs of deep learning relying on big data-driven learning
That is to say, the existing combination of deep neural network and small data is difficult to train, and cannot be well applied to the classification of complexion in traditional Chinese medicine.

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
  • TCM complexion automatic sorting method applying superficial layer neural network
  • TCM complexion automatic sorting method applying superficial layer neural network
  • TCM complexion automatic sorting method applying superficial layer neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] According to the above description, the following is a specific experimental process, but the scope of protection of this patent is not limited to the implementation process, the flow chart is as attached figure 1 shown. The specific implementation process is as follows:.

[0019] Step 1: Construct a training dataset of human facial complexion images.

[0020] Image collection is the basis of facial complexion classification, and the quality of collected images will directly affect the improvement of facial complexion classification accuracy. Under normal circumstances, a darkroom or dark box is the most ideal shooting environment, which can avoid the interference of external stray light and maintain the relative stability of the light source environment.

[0021] Step 1.1: Collect facial complexion images.

[0022] The data collection environment is as follows:

[0023] (1) Use a closed collection environment to avoid stray light entering the shooting environment a...

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

A TCM (Traditional Chinese Medicine) complexion automatic sorting method applying a superficial layer neural network belongs to the field of computer vision. The designed superficial layer neural network has five layers and adopts three different layer structures including an input layer, a characteristic extraction layer and an output layer. The input layer is composed of a convolution layer and a rectified linear unit (ReLU). The characteristic extraction layer is composed of three layers of network, wherein each of the first two layers is composed of a convolution layer and a ReLU activation function; one of the convolution layer and the ReLU is subjected to batch normalization and a pooling layer is added behind the second ReLU of the characteristic extraction layer; the third layer of the characteristic extraction layer is a whole connection layer connected with a ReLU in a followed manner. The output layer is composed of the whole connection layer and is provided with a followed softmax sorter. The method provided by the invention is distinctively advantaged in sorting precision and has invariance property in distortion such as zooming, translation, rotation and the like, is high in robustness and capable of improving sorting precision effectively. Therefore, the depth learning theory is applied to TCM face diagnose objectivity study.

Description

technical field [0001] The present invention takes the human facial image as the research object, on the basis of comprehensively analyzing the characteristics of the human facial image, and utilizes the latest research achievement in the field of artificial intelligencedeep learning technology, to propose a method for automatic classification of complexion in traditional Chinese medicine by applying a shallow neural network , this method automatically learns the deep features of human facial images to classify complexion, avoids the uncertain factors produced by manual feature selection, and improves the accuracy and robustness of facial complexion classification in traditional Chinese medicine. The invention belongs to the field of computer vision and the field of face-to-face diagnosis of traditional Chinese medicine, and specifically relates to technologies such as deep learning and image processing. Background technique [0002] The main basis of TCM diagnosis is the 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
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/00G06T7/11G06K9/62
CPCG06T7/0012G06T7/11G06T2207/20021G06T2207/20081G06T2207/20084G06T2207/30088G06T2207/30201G06T2207/10024G06F18/24G06F18/214
Inventor 张菁肖庆新张辉李晓光卓力
Owner BEIJING UNIV OF 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