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

Millimeter wave radar material identification method based on machine learning

A millimeter-wave radar and machine learning technology, applied in the field of material recognition, can solve the problem that material recognition cannot be realized normally, and achieve the effect of low power consumption, low calculation amount and low cost

Pending Publication Date: 2020-02-07
ZHEJIANG COLLEGE OF ZHEJIANG UNIV OF TECHOLOGY
View PDF6 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the above-mentioned defects in the prior art, the present invention provides a material identification method based on machine learning for millimeter-wave radar, the purpose of which is to provide a material identification method with no damage, low power consumption, low cost, and low complexity , to solve the technical problems in the prior art that the target to be tested of different materials or the target cannot be normally identified in a non-transparent plastic bag, paper bag, plastic box, or carton

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
  • Millimeter wave radar material identification method based on machine learning
  • Millimeter wave radar material identification method based on machine learning
  • Millimeter wave radar material identification method based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] In order to achieve the above-mentioned goal, the data received by the millimeter-wave radar sensor in the present embodiment 1 obtains the machine learning classification model through the training of the naive Bayesian classifier. The optimal category solution is obtained by using the envelope information of the return, and combined with the naive Bayesian classifier, the material of the object to be tested can be more accurately identified.

[0032] Such as figure 1 It mainly includes the following steps:

[0033] Step 1. Use the millimeter wave radar sensor to collect target information;

[0034] It is mainly to receive the electromagnetic wave envelope returned by objects of different materials in the measured three-dimensional fan-shaped area at different distances, record the returned peak data and store it in the database, including the distance data, signal strength, phase and amplitude in the envelope. pair of the same object

[0035] The envelopes generate...

Embodiment 2

[0060] In order to achieve the goal of the present invention, Embodiment 2 implements material identification by combining millimeter-wave radar with a convolutional neural network classifier, including the following steps:

[0061] Step 1. Use the millimeter wave radar sensor to collect target information;

[0062] It is mainly to receive the electromagnetic wave envelope returned by objects of different materials in the measured three-dimensional fan-shaped area at different distances, record the returned peak data and store it in the database, including the distance data, signal strength, phase and amplitude in the envelope.

[0063] Step 2. Data preprocessing: Denoise the electromagnetic wave envelope through the wavelet threshold method, and manually mark the denoised data to generate a sample library;

[0064] Step 2.1, denoise the original electromagnetic wave signal by the wavelet threshold method, and select the soft threshold function as shown in formula 1:

[0065]...

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 millimeter-wave radar material identification method based on machine learning, and relates to the technical field of material identification, and the method comprises the steps: collecting target envelope information through a millimeter-wave radar sensor; de-noising the electromagnetic wave envelope, and manually marking the de-noised data to generate a sample library; constructing a machine learning classifier; measuring the distance between a sensor and an object through millimeter-wave radar, and judging the material of the object by combining a machine learningclassifier. Object materials are marked by detecting millimeter waves to obtain a large number of samples, a sample library is generated, a classifier is trained through a machine learning method, andfinally the object materials are judged in combination with the machine learning classifier and the envelope of electromagnetic waves reflected by a detection object. Compared with an existing identification method, the method has the advantages of being free of damage, low in power consumption, low in cost and low in complexity, and materials of to-be-detected targets or targets made of different materials can be normally identified in non-transparent plastic bags or paper bags or plastic boxes or paper boxes.

Description

technical field [0001] The invention relates to a material identification method, especially a machine learning-based millimeter wave radar material identification method, Background technique [0002] With the rapid development of all walks of life, the demand for identification and detection of object materials in daily life and industrial supervision is getting higher and higher. For example, a sweeping robot needs to be able to distinguish whether the cleaning area is a carpet or a wooden board, so as to determine the best cleaning mode, and to avoid areas with water on the ground, so it is necessary to distinguish the surface material of the cleaned area; for example, the industrial cleaning industry needs to be based on different cleaning materials. And choose different cleaning agents, so the mechanized cleaning process needs to judge the material in advance. Therefore, there is an urgent need for a material detection method that can be integrated into individual lif...

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): G06K9/62G06N3/04G06N3/08G01S13/88
CPCG06N3/08G01S13/88G06N3/045G06F18/24155G06F18/214
Inventor 张智蒋雅晨刘子瑜
Owner ZHEJIANG COLLEGE OF ZHEJIANG UNIV OF TECHOLOGY
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