CN113376609B - Liquid identification method based on millimeter wave radar - Google Patents
Liquid identification method based on millimeter wave radar Download PDFInfo
- Publication number
- CN113376609B CN113376609B CN202110675453.4A CN202110675453A CN113376609B CN 113376609 B CN113376609 B CN 113376609B CN 202110675453 A CN202110675453 A CN 202110675453A CN 113376609 B CN113376609 B CN 113376609B
- Authority
- CN
- China
- Prior art keywords
- liquid
- reflection
- millimeter wave
- information
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/411—Identification of targets based on measurements of radar reflectivity
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N22/00—Investigating or analysing materials by the use of microwaves or radio waves, i.e. electromagnetic waves with a wavelength of one millimetre or more
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/417—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section involving the use of neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Electromagnetism (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Radar Systems Or Details Thereof (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a liquid identification method based on a millimeter wave radar, which comprises the steps of converting a time domain signal sensed by an FMCW millimeter wave radar into a frequency domain signal through fast Fourier change; selecting a peak value area in the frequency domain signal for feature extraction; extracting features by using signals sensed by multiple antennas on a radar, and calculating target information as input information of a neural network; extracting reflection features from the reflection information of the target by using a neural network, and obtaining correction features according to the position information; and utilizing a self-adaptive fusion module to evaluate the interference degree of the change of the target position on the reflected signal according to the extracted characteristics, removing the influence of the position change in a self-adaptive manner, finally obtaining characteristics for distinguishing different liquid components, and predicting the type of the liquid. The method is different from the traditional liquid identification method, does not need to be immersed in liquid and does not need to be in contact with a liquid container, breaks through the limitation of the existing wireless sensing method on identifying the granularity, and can robustly realize the fine-granularity liquid identification effect with high accuracy in different environments.
Description
Technical Field
The invention relates to the technical field of liquid identification, in particular to a liquid identification method based on a millimeter wave radar.
Background
Liquid identification technology has brought a series of significant changes to human life in recent years. It eliminates the need for direct contact with the liquid and indirectly senses the characteristics of the liquid through a machine to identify different liquids. The non-contact identification technology effectively avoids potential safety hazards of toxic liquid to human bodies, and avoids complex operations such as opening of liquid containers and the like.
Fine-grained liquid identification can even go beyond human perception, distinguishing liquids that human senses cannot distinguish: such as white spirit with alcohol concentration of 52 degrees and 53 degrees, fresh milk and expired milk, coca cola and pepertree. Fine particle size liquid identification has numerous important applications in daily life. For example:
1. the liquid safety inspection device is used for detecting flammable and combustible liquid in liquid safety inspection of airports and railway stations;
2. detecting whether the milk is deteriorated in an intelligent supermarket;
3. detecting the change of the blood sugar concentration of a human body in an intelligent health scene;
4. the water quality pollution detection is provided for the war zone;
5. analysis of crude oil water content in oil exploration, etc.
Conventional liquid fine-grain identification typically requires the use of expensive and cumbersome spectrometers to be performed in a laboratory environment. The principle of this type of operation is: different liquids have unique absorption and scattering properties for different frequencies of light, and therefore a spectrometer can be used accordingly for liquid identification. Such methods are limited by instruments and thus are difficult to be widely used in daily life of people.
In order to achieve more convenient liquid identification, many researchers have recently explored the possibility of applying wireless sensing techniques, i.e., analyzing the properties of liquids using the changes that occur when a wireless signal is reflected by or penetrates through the liquid. Compared with the traditional liquid identification technology, the wireless sensing technology has the following advantages: the equipment is cheaper, has better mobility and is easier to deploy in daily life scenes. Although the current wireless sensing technology has been greatly developed, at the stage of just starting, the problems of limited identification range, low identification granularity, low identification effect and the like exist. How to realize the non-contact liquid identification with simpler equipment deployment, thinner identification granularity and more robust effect is still a problem which is widely concerned and needs to be solved.
Disclosure of Invention
The invention aims to provide a liquid identification method based on a millimeter wave radar.
In order to achieve the above purpose, the invention provides the following technical scheme:
a liquid identification method based on a millimeter wave radar comprises the following steps:
s1, converting the time domain signal sensed by the FMCW millimeter wave radar into a frequency domain signal after fast Fourier transform;
s2, selecting a peak area in the frequency domain signal for feature extraction;
s3, extracting features by using signals sensed by multiple antennae on a radar, and calculating target information including position information and reflection information as input information of a neural network;
s4, extracting reflection features from the reflection information of the target by using a neural network, and obtaining correction features according to the position information;
s5, utilizing the self-adaptive fusion module to evaluate the interference degree of the target position change on the reflected signal according to the extracted characteristics, removing the influence of the position change, finally obtaining characteristics for distinguishing different liquid components, and predicting the type of the liquid.
Further, the position information in step S3 includes a target-to-radar distance d, a horizontal angle θ, and a pitch angle β.
Further, the neural network in step S4 includes a correction feature extraction module, a reflection feature extraction module, and an adaptive fusion module, and divides the input vector into position information and reflection information, and the reflection information is sent to the reflection feature extraction module to extract fine-grained reflection features representing the millimeter waves by the target liquid; the position information is sent to a correction feature extraction module that learns the mapping between the target position and the correction features to generate a correction feature representing a correction of the position disturbance.
Furthermore, the correction feature extraction module and the reflection feature extraction module are multilayer perceptrons with five-layer structures, and each layer is composed of a full-connection layer, a regularization layer and an activation function.
Further, the adaptive fusion module automatically selects the correction amplitude according to the magnitude of the characteristic value, specifically, the adaptive fusion module automatically calculates a weight w, weights and fuses the reflection characteristic and the correction characteristic, and finally generates a characteristic F for determining the target categorymThe calculation formula is as follows:
wherein FcRepresenting a corrective characteristic, FrRepresenting the reflection characteristics.
Further, a method of learning using neural networks, based on FcAnd FrThe weight w is calculated according to the magnitude of the characteristic value, and the calculation formula is as follows:
where ψ represents a learnable parameter, exp represents an exponential function with a natural constant e as the base.
Further, the FMCW millimeter wave radar adopts a 60GHz millimeter wave radar.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention relates to a liquid identification method based on a millimeter wave radar, which is a fine-grained sensing method based on the millimeter wave radar, and the method can sense the accurate reflection information and position information of a liquid target by means of the millimeter wave radar, extract the type characteristics of the fine-grained liquid through a customized deep learning neural network model, and can distinguish very similar liquids with high accuracy, such as white spirit with the alcohol concentration difference of only 1 degree, crude oil with different water contents, blood with different blood sugar concentrations, and the like.
2. The liquid identification method based on the millimeter wave radar disclosed by the invention reveals the relation between the slight position change of the liquid container and the signal change on the multiple antennae of the millimeter wave radar. When the position of the liquid target changes, the signal intensity changes on the same side receiving antenna in the antenna array have high consistency. The invention utilizes the correlation of multiple receiving antennas to overcome the interference caused by slight change of the position of the container and obtain more robust liquid type characteristics.
3. The invention discloses a liquid identification method based on a millimeter wave radar, which is a novel multi-antenna array signal processing method. The peak value part (representing the area where the liquid target is located) in the frequency domain signal is selected and then the advantages of multiple receiving antennas of the millimeter wave radar are fully utilized, and accurate target reflection information and position information are calculated.
4. According to the liquid identification method based on the millimeter wave radar, a customized deep learning neural network model is designed, reflection features are extracted from reflection information of a target, and correction features are obtained according to position information and are used for solving signal disturbance caused by position fine transformation. In the network, a self-adaptive fusion module is designed, the degree of position interference can be evaluated according to the extracted characteristics, the influence of position change on a reflected signal is removed, characteristics for distinguishing different liquid materials are obtained finally, and the most possible category of the liquid is predicted.
5. The liquid identification method based on the millimeter wave radar constructs a system prototype by using the millimeter wave radar, the system equipment has small volume (only the size of a coin), is simple to deploy, and can carry out high-precision liquid identification in a non-contact manner, so that the method has wide application scenes: the intelligent monitoring system can be deployed in an intelligent wearable watch to detect blood sugar concentration change, in petroleum exploration equipment to measure crude oil water content, on a shelf of an intelligent supermarket to detect milk deterioration, in a liquid security check scene of an airport to detect flammable and explosive liquid and the like, and has a wide application range.
Drawings
In order to more clearly illustrate the embodiments of the present application or technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings can be obtained by those skilled in the art according to the drawings.
Fig. 1 is a schematic diagram illustrating a principle of a liquid identification method based on a millimeter wave radar according to an embodiment of the present invention.
Fig. 2 is a change curve of RSS after water and white spirit are respectively poured into the cup provided by the embodiment of the present invention, in which curve 1 is water, curve 2 is 53 degrees white spirit, and curve 3 is 75 degrees white spirit.
Fig. 3 is a flowchart of a liquid identification method based on a millimeter wave radar according to an embodiment of the present invention.
Fig. 4 is a schematic structural diagram of a neural network according to an embodiment of the present invention.
Detailed Description
The basic principle of the liquid identification method based on the millimeter wave radar provided by the invention is shown in fig. 1, the FMCW millimeter wave radar can sense the millimeter wave signals reflected by the object, and the Signal Strength (Received Signal Strength, RSS) of the reflected signals is different due to the difference of target characteristics (for example, the liquid dielectric constant is different). The change in RSS after pouring water and white spirit, respectively, into the cup is shown in FIG. 2. The results show that the larger the dielectric constant, the larger the RSS of the reflected millimeter-wave signal. The intensity of millimeter wave signals reflected by the white spirit as non-conductive liquid is obviously less than that of water.
The flow chart of the liquid identification method based on the millimeter wave radar provided by the invention is shown in fig. 3.
Firstly, a time domain signal sensed by an FMCW millimeter wave radar is subjected to fast Fourier transform and converted into a frequency domain signal, then a peak part (representing a region where a liquid target is located) in the frequency domain signal is focused, and the peak part is selected for subsequent feature extraction. And then, the advantages of multiple antennas on the radar are fully utilized, and accurate target information including position information and reflection information is calculated. And finally, the designed neural network (RC-Net) extracts reflection characteristics from the reflection information of the target and obtains correction characteristics according to the position information, so that signal disturbance caused by position fine transformation is solved. A self-adaptive fusion module is designed in the RC-Net, the degree of position interference can be evaluated according to the extracted characteristics, the influence of position change on a reflection signal is removed, characteristics for distinguishing different liquid materials are finally obtained, and the most possible category of the liquid is predicted.
Specifically, regarding a signal processing method of a multi-antenna array, in the past, a time sequence signal collected by a millimeter wave radar is often directly lost to a machine learning method to extract features, and the coarse mode is often difficult to learn fine-grained target features. The invention firstly uses fast Fourier transform to convert an Intermediate Frequency (Intermediate Frequency) signal obtained by FMCW radar processing from a time domain signal to a Frequency domain signal. The peak area on the frequency domain signal of the FMCW radar represents the area where the target is located. And carrying out feature extraction only on the target area to focus the target and extract fine-grained features of the target liquid. For the target area, except the reflection intensity R of the signals sensed by the multiple receiving antennas, the invention fully utilizes the advantages of the multiple receiving antennas of the millimeter wave radar to calculate and obtain detailed target position information: the distance d from the target to the radar, the horizontal angle theta and the pitch angle beta (i.e., the arrival angle of the signal in the horizontal direction and the arrival angle of the signal in the pitch direction). And finally, the obtained group of target information is used as input information of the RC-Net.
With respect to the neural network RC-Net for fluid identification, as shown in fig. 4, RC-Net has three modules: the device comprises a correction feature extraction module, a reflection feature extraction module and a self-adaptive fusion module. The input vector is divided into position information and reflection information, and the reflection information is sent to a reflection feature extraction module to be used for extracting fine-grained reflection features representing the millimeter waves of target liquid. The position information is sent to a correction feature extraction module, which learns the mapping relationship between the target position and the correction features to generate a correction feature representing the correction of the disturbance caused by the position change.
The two feature extraction modules have similar structures, and each layer of the multilayer perceptron with a 5-layer structure is composed of a fully-connected layer, a regularization layer and an activation function (ReLU). The number of input and output eigen-channels per layer is shown in fig. 4. N =3 in fig. 4, indicating that three data points were selected for the target area.
Considering that the reflection characteristics are different in the degree of position interference and different in the required correction degree, a self-adaptive fusion module is designed, and the correction amplitude is automatically selected according to the size of the characteristic value. Specifically, the module automatically calculates a weight w, weights and fuses the reflection feature and the correction feature, and finally generates a feature F for determining the target classm. The calculation formula is as follows:
wherein FcRepresenting a corrective characteristic, FrRepresenting the reflection characteristics.
The invention uses a neural network learning method based on FcAnd FrThe weight w is calculated according to the magnitude of the characteristic value, and the calculation formula is as follows:
where ψ represents a learnable parameter, exp represents an exponential function with a natural constant e as the base.
Specifically, we use two fully-connected layers to compute the value of the weight, i.e., ψ represents a parameter that can be learned in the two fully-connected layers. The number of neurons in the fully connected layer was 64 and 1, respectively.
The method breaks through the limitation of the existing wireless sensing method on identifying the granularity, and can robustly realize the effect of identifying the fine-grained liquid with high accuracy in different environments. As a non-contact liquid identification scheme, namely the liquid identification scheme does not need to be immersed in liquid and does not need to be in contact with a liquid container, the technical scheme of the invention provides a portable high-accuracy liquid identification method for a plurality of important liquid identification scenes in daily life.
The method can be deployed and used in the following scenes: 1) in the liquid security inspection of airports and railway stations, whether flammable and combustible liquid exists in a container carried by passengers is detected; 2) the milk quality monitoring system is deployed on a commodity shelf in an intelligent supermarket and used for dynamically monitoring whether milk is deteriorated or not; 3) the method comprises the steps of detecting human blood glucose concentration changes in wearable equipment deployed in an intelligent health scene; 4) the method provides detection of water pollution for the war zone and analyzes whether the water source is polluted; 5) analysis of crude oil water content in oil exploration, etc.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: it is to be understood that modifications may be made to the technical solutions described in the foregoing embodiments, or equivalents may be substituted for some of the technical features thereof, but such modifications or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (4)
1. A liquid identification method based on a millimeter wave radar is characterized by comprising the following steps:
s1, converting the time domain signal sensed by the FMCW millimeter wave radar into a frequency domain signal after fast Fourier transform;
s2, selecting a peak area in the frequency domain signal for feature extraction;
s3, extracting features by using signals sensed by multiple antennae on a radar, and calculating target information including position information and reflection information as input information of a neural network;
s4, extracting reflection features from the reflection information of the target by using a neural network, and obtaining correction features according to the position information;
the neural network in the step S4 includes a correction feature extraction module, a reflection feature extraction module, and an adaptive fusion module, and divides the input vector into position information and reflection information, and the reflection information is sent to the reflection feature extraction module to extract fine-grained reflection features representing the millimeter waves by the target liquid; the position information is sent to a correction feature extraction module, and the module learns the mapping relation between the target position and the correction feature to generate a correction feature representing the correction of the position disturbance;
s5, utilizing the self-adaptive fusion module to evaluate the interference degree of the target position change on the reflected signal according to the extracted characteristics, removing the influence of the position change, finally obtaining characteristics for distinguishing different liquid components, and predicting the type of the liquid.
2. The millimeter wave radar-based liquid identification method according to claim 1, wherein the correction feature extraction module and the reflection feature extraction module are multilayer perceptrons having a five-layer structure, and each layer is composed of a full connection layer, a regularization layer and an activation function.
3. The millimeter wave radar-based liquid identification method according to claim 1, wherein the adaptive fusion module automatically selects the correction amplitude according to the magnitude of the characteristic value, and specifically, the adaptive fusion module automatically calculates a weight w, weights and fuses the reflection characteristic and the correction characteristic, and finally generates a characteristic F for determining the target categorymThe calculation formula is as follows:
Fm=w*Fc+(1-w)*Fr
wherein FcRepresenting a corrective characteristic, FrRepresenting the reflection characteristics.
4. The millimeter wave radar-based liquid identification method according to claim 3, wherein the method of neural network learning is based on FcAnd FrThe weight w is calculated according to the magnitude of the characteristic value, and the calculation formula is as follows:
where ψ represents a learnable parameter, exp represents an exponential function with a natural constant e as the base.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110675453.4A CN113376609B (en) | 2021-06-18 | 2021-06-18 | Liquid identification method based on millimeter wave radar |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110675453.4A CN113376609B (en) | 2021-06-18 | 2021-06-18 | Liquid identification method based on millimeter wave radar |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113376609A CN113376609A (en) | 2021-09-10 |
CN113376609B true CN113376609B (en) | 2021-12-31 |
Family
ID=77577690
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110675453.4A Active CN113376609B (en) | 2021-06-18 | 2021-06-18 | Liquid identification method based on millimeter wave radar |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113376609B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114019496B (en) * | 2022-01-05 | 2022-03-08 | 北京邮电大学 | Non-contact measurement method and device for flow velocity of liquid in pipeline |
CN117158940A (en) * | 2022-09-01 | 2023-12-05 | 亿慧云智能科技(深圳)股份有限公司 | Heart rate data monitoring method, detection system and wearable device using millimeter wave radar |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103630555A (en) * | 2013-11-21 | 2014-03-12 | 烟台大学 | Method for detecting liquid hazardous substances by adopting microwave multi-frequency-point vector detection |
CN109470720A (en) * | 2018-09-29 | 2019-03-15 | 深圳市华讯方舟太赫兹科技有限公司 | Liquid identification method, vector extracting method, Liquid identification device and storage medium |
CN110245675A (en) * | 2019-04-03 | 2019-09-17 | 复旦大学 | A kind of dangerous objects detection method based on millimeter-wave image human body contextual information |
CN110361071A (en) * | 2018-04-11 | 2019-10-22 | 英飞凌科技股份有限公司 | The device and method for carrying out liquid detecting using millimetre-wave radar sensor |
CN112130143A (en) * | 2019-06-25 | 2020-12-25 | 富士通株式会社 | Article detection method and apparatus |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9429528B2 (en) * | 2013-05-23 | 2016-08-30 | Texas Instruments Incorporated | Determining gas absorption line from separate and alternating RF signals |
-
2021
- 2021-06-18 CN CN202110675453.4A patent/CN113376609B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103630555A (en) * | 2013-11-21 | 2014-03-12 | 烟台大学 | Method for detecting liquid hazardous substances by adopting microwave multi-frequency-point vector detection |
CN110361071A (en) * | 2018-04-11 | 2019-10-22 | 英飞凌科技股份有限公司 | The device and method for carrying out liquid detecting using millimetre-wave radar sensor |
CN109470720A (en) * | 2018-09-29 | 2019-03-15 | 深圳市华讯方舟太赫兹科技有限公司 | Liquid identification method, vector extracting method, Liquid identification device and storage medium |
CN110245675A (en) * | 2019-04-03 | 2019-09-17 | 复旦大学 | A kind of dangerous objects detection method based on millimeter-wave image human body contextual information |
CN112130143A (en) * | 2019-06-25 | 2020-12-25 | 富士通株式会社 | Article detection method and apparatus |
Non-Patent Citations (2)
Title |
---|
Detection of Ammonia in Liquids Using Millimeter Wave Spectroscopy;Ozturk, Hilmi等;《INTERNATIONAL JOURNAL OF PHOTOENERGY》;20121231;第2012卷;1-4页 * |
无线感知液体的紧凑型无芯RFID标签;陈娜等;《2019年全国天线年会论文集》;20190722;1069-1972页 * |
Also Published As
Publication number | Publication date |
---|---|
CN113376609A (en) | 2021-09-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113376609B (en) | Liquid identification method based on millimeter wave radar | |
Geng et al. | Multi-scale deep feature learning network with bilateral filtering for SAR image classification | |
CN102799903B (en) | High-spectrum automatic cloud detection method based on space and spectral information | |
Liang et al. | FG-LiquID: A contact-less fine-grained liquid identifier by pushing the limits of millimeter-wave sensing | |
CN110427878A (en) | A kind of sudden and violent signal recognition method of Rapid Radio and system | |
CN113391282A (en) | Human body posture recognition method based on radar multi-dimensional feature fusion | |
CN113837131A (en) | Multi-scale feature fusion gesture recognition method based on FMCW millimeter wave radar | |
CN114139566A (en) | Method for improving accuracy of measurement based on machine learning weighted average fusion feature extraction | |
Yang et al. | On the feasibility of estimating soluble sugar content using millimeter-wave | |
Qu et al. | Dynamic hand gesture classification based on multichannel radar using multistream fusion 1-D convolutional neural network | |
Soumya et al. | Recent advances in mmWave-radar-based sensing, its applications, and machine learning techniques: A review | |
Sim et al. | Road structure classification through artificial neural network for automotive radar systems | |
CN110632132A (en) | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network | |
Rather et al. | Evaluation of machine learning models for a chipless RFID sensor tag | |
Mabu et al. | Anomaly Detection Using Convolutional Adversarial Autoencoder and One-class SVM for Landslide Area Detection from Synthetic Aperture Radar Images. | |
Hoog et al. | A 60 GHz pulsed coherent radar for online monitoring of the withering condition of leaves | |
Chuang et al. | Applying an Adaptive Signal Identification Method to Improve Vessel Echo Detection and Tracking for SeaSonde HF Radar | |
Yu et al. | CNN-based weather signal detection algorithm for airborne weather radar | |
CN113093304B (en) | Suspicious article material safety detection method based on WIFI | |
Hwang et al. | Multifunctional scatterometer system for measuring physical oceanographic parameters using range-Doppler FMCW radar | |
Zhang et al. | RF-LqRNN: RFID-based concentration detection of alcohol solutions and glucose solutions | |
Chen et al. | Convolution neural network and 77 GHz millimeter wave radar based intelligent liquid classification system | |
Wilson et al. | Multitarget angle of arrival estimation using rotating mmwave fmcw radar and yolov3 | |
CN117056794A (en) | Non-contact liquid component recognition model training method, recognition method, system and device | |
Hu et al. | mm-Wheat: Non-contact Detection of Wheat Moisture Content with Millimeter-Wave Radar |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |