WO2021068470A1 - 一种基于雷达信号的身份及手势识别方法 - Google Patents

一种基于雷达信号的身份及手势识别方法 Download PDF

Info

Publication number
WO2021068470A1
WO2021068470A1 PCT/CN2020/083989 CN2020083989W WO2021068470A1 WO 2021068470 A1 WO2021068470 A1 WO 2021068470A1 CN 2020083989 W CN2020083989 W CN 2020083989W WO 2021068470 A1 WO2021068470 A1 WO 2021068470A1
Authority
WO
WIPO (PCT)
Prior art keywords
gesture
radar
signal
identity
information
Prior art date
Application number
PCT/CN2020/083989
Other languages
English (en)
French (fr)
Inventor
王勇
陈君毅
曹佳禾
Original Assignee
浙江大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 浙江大学 filed Critical 浙江大学
Priority to PCT/CN2020/083989 priority Critical patent/WO2021068470A1/zh
Priority to US17/218,174 priority patent/US11947002B2/en
Publication of WO2021068470A1 publication Critical patent/WO2021068470A1/zh

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
    • 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
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • 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
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/06—Systems determining position data of a target
    • G01S13/08—Systems for measuring distance only
    • G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
    • G01S13/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
    • G01S13/343—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal using sawtooth modulation
    • 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
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/50—Systems of measurement based on relative movement of target
    • G01S13/505—Systems of measurement based on relative movement of target using Doppler effect for determining closest range to a target or corresponding time, e.g. miss-distance indicator
    • 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
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/50—Systems of measurement based on relative movement of target
    • G01S13/52—Discriminating between fixed and moving objects or between objects moving at different speeds
    • G01S13/522—Discriminating between fixed and moving objects or between objects moving at different speeds using transmissions of interrupted pulse modulated waves
    • G01S13/524—Discriminating between fixed and moving objects or between objects moving at different speeds using transmissions of interrupted pulse modulated waves based upon the phase or frequency shift resulting from movement of objects, with reference to the transmitted signals, e.g. coherent MTi
    • G01S13/53—Discriminating between fixed and moving objects or between objects moving at different speeds using transmissions of interrupted pulse modulated waves based upon the phase or frequency shift resulting from movement of objects, with reference to the transmitted signals, e.g. coherent MTi performing filtering on a single spectral line and associated with one or more range gates with a phase detector or a frequency mixer to extract the Doppler information, e.g. pulse Doppler radar
    • 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
    • G01S7/415—Identification of targets based on measurements of movement associated with the target
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/017—Gesture based interaction, e.g. based on a set of recognized hand gestures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M1/00—Substation equipment, e.g. for use by subscribers
    • H04M1/02—Constructional features of telephone sets
    • H04M1/0202—Portable telephone sets, e.g. cordless phones, mobile phones or bar type handsets
    • H04M1/026—Details of the structure or mounting of specific components
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M1/00—Substation equipment, e.g. for use by subscribers
    • H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
    • H04M1/724—User interfaces specially adapted for cordless or mobile telephones
    • H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
    • H04M1/72454—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions according to context-related or environment-related conditions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M1/00—Substation equipment, e.g. for use by subscribers
    • H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
    • H04M1/724—User interfaces specially adapted for cordless or mobile telephones
    • H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
    • H04M1/72463—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions to restrict the functionality of the device
    • H04M1/724631—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions to restrict the functionality of the device by limiting the access to the user interface, e.g. locking a touch-screen or a keypad

Definitions

  • the invention relates to the field of identity recognition, in particular to an identity and gesture recognition method based on radar signals.
  • Gesture recognition refers to the entire process of tracking human gestures, recognizing their representations, and converting them into semantically meaningful commands. As an important part of human-computer interaction, its research and development affects the naturalness and flexibility of human-computer interaction. With the development of technology at this stage, people use gesture recognition applications on various devices increasing day by day.
  • gesture recognition usually uses a camera to collect gesture information, and there is a certain degree of waste in energy consumption.
  • the amount of information contained in gesture images is huge, and it is often prone to privacy issues such as information leakage.
  • the currently used gesture recognition often requires an additional authentication method to control whether the device is unlocked before the device is unlocked. After the device is successfully unlocked, the gesture recognition can be continued, which is relatively cumbersome.
  • the device is unlocked, anyone who makes the same gesture on the device can produce the same effect. This situation has certain risks.
  • the purpose of the present invention is to provide an identity and gesture recognition method based on radar signals, which can solve the need for additional unlocking steps in gesture recognition and a technology that can produce the same effect when anyone makes the same gesture on the device after the device is unlocked. problem.
  • an identity and gesture recognition method based on radar signals including the following steps:
  • Step 1 Read the radar echo signal received by the radar sensor, which is reflected by various gestures of the tester;
  • Step 2 Mix the radar echo signal with the radar transmission signal
  • Step 3 Use a high-pass filter to filter the mixed signal
  • Step 4 Centralize the filtered signal data
  • Step 5 Design a neural network model suitable for radar signal characteristics, use the signal preprocessed in step 4 and the tester’s identity information label to train the model to obtain a neural network A capable of identity verification;
  • Step 6 According to the preprocessed signal and radar sensor parameters in step 4, the distance, speed and angle information are obtained, so as to calculate the three-dimensional coordinates and Doppler value of the moving target;
  • Step 7 Use the space grid method to divide the space above the radar; map the space area into a three-dimensional matrix, and add the Doppler value of the three-dimensional coordinates and Doppler value obtained in step 6 to the grid position in the space to which they belong
  • the grid corresponds to the size of the matrix unit element in the matrix unit; send the matrix and gesture information tags into the designed neural network to train to obtain a neural network B capable of gesture recognition;
  • Step 8 Perform constant false alarm detection on the radar echo signal detected in real time to determine whether someone is performing gesture recognition, if yes, proceed to step 9; if no, continue to wait for detection;
  • Step 9 Preprocess the detected radar echo signal from step 2 to step 4; send the processed signal to the neural network A trained in step 5 to extract features for analysis; verify the identity information contained in the feature The user's identity, if the identity verification is passed, the user's gesture is valid, go to step 10, otherwise continue to wait for detection;
  • Step 10 After the data preprocessed in steps 2 to 4 are processed in steps 6 and 7, input the neural network B to determine the user's gesture, and perform corresponding operations according to the correspondence between the user's gesture and the operation.
  • the radar sensor adopts an FMCW millimeter wave radar with a frequency range of 57.4 GHz to 62.6 GHz.
  • the multiple corresponding waveforms obtained are averaged and then filtered.
  • an eighth-order Butterworth high-pass filter with a cut-off frequency of 31250 Hz is used to filter the mixed signal to filter out large DC and low-frequency noise while retaining effective information.
  • the distance information R is obtained through the fast time dimension FFT:
  • f movingBeat and f staticBeat are the frequency of the beat signal in the moving and stationary state of the target
  • f d is the Doppler frequency
  • f c is the sweep bandwidth
  • R is the target distance
  • C is the speed of light
  • t c is the sweep period
  • F is the center frequency of the Chirp signal
  • v is the target speed
  • ⁇ d is the distance difference between any two receiving antennas and the target
  • ⁇ is the phase difference of the signals received by the corresponding two antennas
  • L is the distance between the corresponding two antennas
  • ⁇ is the wavelength
  • the area with a spatial size of Xcm*Ycm*Zcm is mapped into a three-dimensional matrix with a size of (X*k1)*(Y*k2)*(Z*k3), where X , Y, Z are the length, width and height corresponding to the space, and k1, k2, and k3 are the mapping coefficients of the actual space length, width and height to the three-dimensional matrix.
  • the effective user establishes the corresponding relationship between the prescribed gesture and the operation during training.
  • the specific gesture of the valid user is recognized as a prescribed gesture in step 10
  • the operation corresponding to the prescribed gesture is executed. If the recognized gesture is not present In the prescribed gesture, no operation is generated.
  • the method is applied to a mobile phone.
  • the mobile phone has a built-in radar sensor and pre-stores the identity information of the device owner and gesture information, and some of the gesture information corresponds to App operation instructions; on the premise that the mobile phone is locked, if gestures are used for identity verification If the person is the owner of the device, the phone is unlocked, the gesture information is determined, and the app corresponding to the gesture is opened. If there is no corresponding gesture, no operation is performed; if the gesture does not belong to the device owner, the phone remains locked; when the phone is unlocked , If the gesture does not belong to the device owner or the gesture made by the device owner is not in the prescribed gestures, the corresponding operation will not be performed.
  • the present invention has the beneficial effects: the radar signal-based identity and gesture recognition method proposed by the present invention uses radar to collect signals, effectively reducing energy consumption and ensuring user privacy. On the premise of performing only one gesture operation, it is possible to determine whether to perform the operation corresponding to the gesture according to whether the gesture belongs to the corresponding user. This method greatly improves the convenience and safety of gesture recognition.
  • Figure 1 is an offline training flow chart of identity and gesture recognition based on radar signals
  • Figure 2 shows how the millimeter wave radar sensor is used
  • Figure 3 is a flowchart of online recognition of identity and gesture recognition based on radar signals.
  • this embodiment provides an identity and gesture recognition method based on radar signals.
  • the FMCW millimeter-wave radar with a frequency range of 57.4GHz to 62.6GHz is used, and the frame rate of the transmitted signal is 60 frames per second, and each frame of data is divided into 8 Chirp signals. A total of 21 people were subjected to this radar signal-based identity and gesture recognition method.
  • the offline training process is shown in Figure 1, and mainly includes the following steps:
  • Step 1 Place the radar sensor on a flat surface.
  • the tester performs the prescribed N gestures and interference gestures (interference gestures, that is, gestures other than N gestures) above the radar sensor in the manner shown in Figure 2, and reads the radar.
  • the radar echo signal received by the sensor through the reflection of various gestures is denoted as S 1 ;
  • Step 2 Mix the radar echo signal S 1 with the radar transmit signal S 2 and the signal obtained after mixing
  • ⁇ 1 and ⁇ 2 represent the frequencies of the transmitted signal and the signal reflected by the palm, respectively, with Denote the phases of the two signals respectively, and mark the mixed signal as D.
  • the millimeter-wave radar used in this example has multiple receiving channels, and multiple corresponding waveforms can be obtained. Take measures to average them, and record the obtained signal as D′;
  • Step 3 Use an eighth-order Butterworth high-pass filter with a cut-off frequency of 31250Hz to filter the signal D′ to obtain signal data M.
  • the selection of this filter is mainly to filter out large DC and low-frequency noise while retaining effective information .
  • Step 5 Design a neural network model suitable for the characteristics of the radar signal (the input layer of the neural network needs to be adapted to the parameters of the collected data preset by the radar), and the above-mentioned preprocessed signal M′ with the identity information label of the tester Put it into it for training, and get a neural network A that can perform identity verification;
  • Step 6 According to M′ and radar sensor parameters, obtain distance information R through fast time dimension FFT:
  • f movingBeat and f staticBeat are the frequency of the beat signal in the moving and stationary state of the target
  • f d is the Doppler frequency
  • f c is the sweep bandwidth
  • R is the target distance
  • C is the speed of light
  • t c is the sweep period
  • F is the center frequency of the Chirp signal
  • v is the target speed.
  • ⁇ d is the distance difference between any two receiving antennas and the target
  • ⁇ is the phase difference of the signals received by the corresponding two antennas
  • L is the distance between the corresponding two antennas
  • ⁇ is the wavelength.
  • Step 7 Use the spatial grid method to divide the space above the radar.
  • the area with the space size of Xcm*Ycm*Zcm is mapped into a three-dimensional matrix of size (X*k1)*(Y*k2)*(Z*k3).
  • X, Y, Z are the length, width and height corresponding to the space
  • k1, k2, and k3 are the mapping coefficients of the actual space length, width and height and the three-dimensional matrix.
  • a 20cm*20cm*10cm actual space can be selected and mapped to a 20*20*10 three-dimensional matrix (that is, a 1cm 3 space grid corresponds to an element unit in the matrix).
  • the data (three-dimensional coordinates and corresponding Doppler value) obtained in step 6 is added to the corresponding matrix unit of the grid as the size of the matrix unit element according to the grid position in the space where it belongs.
  • the online identification process is shown in Figure 3, and it mainly includes the following steps:
  • Step 8 Perform constant false alarm detection on the detected radar echo signal to determine whether someone is performing gesture recognition, if yes, proceed to step 9; if no, continue to wait for detection;
  • Step 9 Perform the preprocessing of the above steps 2 to 4 on the detected radar echo signal; send the processed signal to the neural network A trained in step 5 to extract features for analysis; use the identity information contained in the features Verify the identity of the user. If the identity verification is passed (the classification of the information by the neural network is the category of the valid user we identified), the user's gesture is valid, proceed to step 10, otherwise continue to wait for detection;
  • Step 10 The data preprocessed from steps 2 to 4 above are processed in steps 6 and 7, and then sent to neural network B to determine which gesture the user has made, so as to perform subsequent corresponding operations. If the gesture is not among the pre-defined N kinds of gestures, no operation will be generated.
  • the mobile phone has a built-in radar sensor, and pre-stores the identity information of the device owner and gesture information, some of the gesture information corresponds to App operation instructions; on the premise that the mobile phone is locked, if the person making the gesture for identity verification is the device owner, the mobile phone is unlocked , Determine the gesture information, open the App corresponding to the gesture, if there is no corresponding gesture, no operation is performed; if the gesture does not belong to the device owner, the phone remains locked.
  • Identity verification and gesture recognition can work at the same time through one gesture. This completes the entire identity and gesture recognition process based on radar signals. After many experiments, it is proved that the present invention can achieve a recognition accuracy of about 90%.
  • the radar signal-based identity and gesture recognition method uses radar to collect signals, effectively reducing energy consumption and ensuring user privacy.
  • the neural network is used to extract the features, and it can be judged whether to perform the operation corresponding to the gesture according to whether the gesture belongs to the corresponding user. This method improves the security of gesture recognition. The entire process does not require additional unlocking steps, and only requires one gesture operation, which also makes gesture recognition more convenient.

Landscapes

  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computer Security & Cryptography (AREA)
  • Software Systems (AREA)
  • Human Computer Interaction (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Signal Processing (AREA)
  • Computer Hardware Design (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Electromagnetism (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Environmental & Geological Engineering (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

一种基于雷达信号的身份及手势识别方法，读取雷达传感器接收到的经由测试人员各种手势反射得到的雷达回波信号（S1）；将雷达回波信号与雷达发射信号进行混频（D）；对混频后的信号进行滤波（M）、中心化操作（M'）后，分别训练得到能够进行身份验证的神经网络（A）和能够进行手势识别的神经网络（B）。在实时检测过程中，首先验证使用者身份，若身份验证通过则该使用者的手势有效，验证使用者手势，根据使用者手势与操作的对应关系进行相应操作。该方法仅通过一次手势操作即可在进行手势识别的同时进行身份验证，能够根据手势是否属于相应用户，判断是否进行手势对应的操作，提高了手势识别的便捷性和安全性。

Description

一种基于雷达信号的身份及手势识别方法 技术领域
本发明涉及身份识别领域，尤其涉及一种基于雷达信号的身份及手势识别方法。
背景技术
手势识别指的是跟踪人类手势、识别其表示和转换为语义上有意义的命令的整个过程。作为人机交互的重要组成部分，其研究发展影响着人机交互的自然性和灵活性。而随着现阶段技术的发展，人们在各种设备上使用手势识别的应用日益增加。
传统的手势识别通常是采用摄像头采集手势信息，在能耗上存在一定程度的浪费，手势图像包含的信息量巨大也往往容易存在信息泄露等隐私方面的问题。而且目前所采用的手势识别，在设备没有解锁之前，往往需要通过额外的身份验证方法来控制设备是否解锁，将设备成功解锁后，才能继续进行手势识别，这一步骤相对繁琐。而且当设备解锁后，任何人对设备做相同的手势都能产生相同的效果，这一情况存在一定的风险性。
因此，如何在设备上进行手势识别的同时进行身份验证，提高手势识别的便捷性和安全性，是当下急需解决的问题。
发明内容
有鉴于此，本发明的目的在于提供一种基于雷达信号的身份及手势识别方法，可以解决手势识别中需要额外的解锁步骤以及设备解锁后任何人对设备做相同手势都能产生相同效果的技术问题。
为解决上述技术问题，本发明提出如下技术方案：一种基于雷达信号的身份及手势识别方法，包括以下步骤：
步骤1、读取雷达传感器接收到的经由测试人员各种手势反射得到的雷达回波信号；
步骤2、将雷达回波信号与雷达发射信号进行混频；
步骤3、采用高通滤波器对混频后的信号进行滤波；
步骤4、将滤波后的信号数据进行中心化操作；
步骤5、设计适用于雷达信号特征的神经网络模型，利用步骤4预处理后的信号与测试人员的身份信息标签对模型进行训练，得到能够进行身份验证的神经网络A；
步骤6、根据步骤4预处理后的信号及雷达传感器参数，得到距离、速度及角度信息，从而计算得到运动目标的三维坐标及Doppler值；
步骤7、采用空间栅格法对雷达上方空间进行划分；将空间区域映射为三维矩阵，将步骤6得到的三维坐标及Doppler值根据其所属空间中的栅格位置，将其Doppler值累加在该栅格对应矩阵单元中作为矩阵单元元素的大小；将该矩阵与手势信息标签送入设计好的神经网络中训练得到能够进行手势识别的神经网络B；
步骤8、将实时检测到的雷达回波信号进行恒虚警检测判断是否有人在进行手势识别，如果有则进行步骤9；无则继续等待检测；
步骤9、将检测到的雷达回波信号进行步骤2到步骤4的预处理；将处理后的信号送入步骤5中训练好的神经网络A提取特征进行分析；通过特征中含有的身份信息验证使用者的身份，若身份验证通过则该使用者的手势有效，进行步骤10，否则继续等待检测；
步骤10、将进行步骤2到步骤4预处理后的数据再进行步骤6与步骤7处理后，输入神经网络B判断使用者手势，根据使用者手势与操作的对应关系进行相应操作。
进一步地，所述雷达传感器采用频率区间为57.4GHz至62.6GHz的FMCW毫米波雷达。
进一步地，在所述步骤2中，若雷达传感器具有多个接收通道，则将得到的多条对应波形取平均之后再进行滤波。
进一步地，在所述步骤3中，采用截止频率为31250Hz的八阶巴特沃斯高通滤波器对混频后的信号进行滤波，在保留有效信息的同时滤除较大的直流和低频噪声。
进一步地，在所述步骤6中，根据步骤4预处理后的信号及雷达传感器参数，通过快时间维度FFT得到距离信息R：
Figure PCTCN2020083989-appb-000001
Figure PCTCN2020083989-appb-000002
再通过慢时间维度FFT得到速度信息v：
Figure PCTCN2020083989-appb-000003
Figure PCTCN2020083989-appb-000004
其中f movingBeat和f staticBeat分别为目标运动和静止状态下差拍信号的频率，f d为多普勒频率，f c为扫频带宽，R为目标距离，C为光速，t c为扫频周期，f为Chirp信号的中心频率，v为目标速度；
根据雷达传感器的多个收发天线得到角度信息θ：
Figure PCTCN2020083989-appb-000005
Figure PCTCN2020083989-appb-000006
其中Δd为任意两个接收天线距离目标的距离差值，ΔΦ为对应两个天线接收信号的相位差，L是对应两个天线之间的距离，λ为波长。
进一步地，在所述步骤7中，将空间大小为Xcm*Ycm*Zcm的区域，映射为一个大小为(X*k1)*(Y*k2)*(Z*k3)的三维矩阵，其中X，Y，Z为空间对应的长宽高，k1，k2，k3为实际空间长宽高与三维矩阵的映射系数。
进一步地，有效使用者在训练时建立规定手势与操作的对应关系，在步骤10识别出有效使用者的具体手势为某一规定手势时，执行该规定手势对应的操作，若识别出的手势不在规定手势中，则不产生操作。
进一步地，该方法应用于手机中，手机内置有雷达传感器并预先存储设备所有者身份信息及手势信息，部分手势信息对应App操作指令；在手机锁屏的前提下，若做手势进行身份验证的人员为设备所有者，则手机解锁，判断手势信息，打开手势对应App，若无对应手势则不进行任何操作；若手势不属于设备所有者，则手机维持锁屏；在手机已经解锁的情况下，若手势不属于设备所有者或设备所有者所做手势不在规定手势中，则不会进行对应的操作。
本发明所具有的有益效果：本发明提出的基于雷达信号的身份及手势识别方法，通过雷达进行信号采集，有效减少能耗，保障用户的隐私。在仅进行一次手势操作的前提下，能够根据手势是否属于相应用户，判断是否进行手势对应的操作。该方法大大提升了手势识别的便捷性和安全性。
附图说明
图1是基于雷达信号的身份及手势识别离线训练流程图；
图2是毫米波雷达传感器的使用方式；
图3是基于雷达信号的身份及手势识别在线识别流程图。
具体实施方式
以下结合附图对本发明实施例的具体实施方式进行详细说明。应当理解的是，此处所描述的具体实施方式仅用于说明和解释本发明实施例，并不用于限制本发明实施例。
不失一般性，本实施实例提供一种基于雷达信号的身份及手势识别方法。采用频率区间为57.4GHz至62.6GHz的FMCW毫米波雷达，发射信号的帧率60帧/秒，每帧数据分为8 个Chirp信号。总计对21个人进行该基于雷达信号的身份及手势识别方法。
离线训练部分流程按图1所示，主要包括以下步骤：
步骤1、将雷达传感器放置于一个平面上，测试人员按照图2的方式在雷达传感器的上方做所规定的N种手势以及干扰手势(干扰手势即N种手势之外的手势)，读取雷达传感器接收到的经由各种手势反射得到的雷达回波信号，记为S 1；
步骤2、将雷达回波信号S 1与雷达发射信号S 2进行混频，混频后得到的信号
Figure PCTCN2020083989-appb-000007
Figure PCTCN2020083989-appb-000008
其中ω 1和ω 2分别表示发射信号和经由手掌反射信号的频率，
Figure PCTCN2020083989-appb-000009
和
Figure PCTCN2020083989-appb-000010
分别表示两个信号的相位，将混频后的信号记为D。本示例采用的毫米波雷达具有多个接收通道，可以得到多条对应波形，对其采取取平均的措施，将得到的信号记为D′；
步骤3、采用截止频率为31250Hz的八阶巴特沃斯高通滤波器对信号D′进行滤波得到信号数据M，该滤波器的选取主要是在保留有效信息的同时滤除较大的直流和低频噪声。
步骤4、将滤波后的信号数据进行中心化操作M′＝M-μ，数据的平均值
Figure PCTCN2020083989-appb-000011
其中n为数据量的大小。
步骤5、设计适用于雷达信号特征的神经网络模型(该神经网络的输入层需要与雷达预设的采集数据的参数适配)，将上述预处理后的信号M′与测试人员的身份信息标签对应放入其中进行训练，得到能够进行身份验证的神经网络A；
步骤6、根据M′及雷达传感器参数，通过快时间维度FFT得到距离信息R：
Figure PCTCN2020083989-appb-000012
Figure PCTCN2020083989-appb-000013
再通过慢时间维度FFT得到速度信息v：
Figure PCTCN2020083989-appb-000014
Figure PCTCN2020083989-appb-000015
其中f movingBeat和f staticBeat分别为目标运动和静止状态下差拍信号的频率，f d为多普勒频率，f c为扫频带宽，R为目标距离，C为光速，t c为扫频周期，f为Chirp信号的中心频率，v为目标速度。
根据雷达传感器的多个收发天线得到角度信息θ：
Figure PCTCN2020083989-appb-000016
Figure PCTCN2020083989-appb-000017
其中Δd为任意两个接收天线距离目标的距离差值，ΔΦ为对应两个天线接收信号的相位差，L是对应两个天线之间的距离，λ为波长。最终根据距离信息R、速度信息v及角度信息θ，推出运动目标的三维坐标(x,y,z)及Doppler值；
步骤7、采用空间栅格法对雷达上方空间进行划分。
将空间大小为Xcm*Ycm*Zcm的区域，映射为一个大小为(X*k1)*(Y*k2)*(Z*k3)的三维矩阵。其中X，Y，Z为空间对应的长宽高，k1，k2，k3为实际空间长宽高与三维矩阵的映射系数。具体可用选取一个20cm*20cm*10cm的实际空间并映射为一个20*20*10的三维矩阵(即1cm 3空间栅格对应矩阵中一个元素单元)。
将步骤6得到的数据(三维坐标及对应Doppler值)根据其所属空间中的栅格位置，将其Doppler值累加在该栅格对应矩阵单元中作为矩阵单元元素的大小。将该矩阵S与手势信息标签对应送入设计好的神经网络中得到能够进行手势识别的神经网络B；
在线识别部分流程如图3所示，主要包括如下步骤：
步骤8、将检测到的雷达回波信号进行恒虚警检测判断是否有人在进行手势识别，如果有则进行步骤9；无则继续等待检测；
步骤9、将检测到的雷达回波信号进行上述步骤2到步骤4的预处理；将处理后的信号送入步骤5中训练好的神经网络A提取特征进行分析；通过特征中含有的身份信息验证使用者的身份，若身份验证通过(神经网络对该信息的分类结果是我们认定的有效使用者的类别)，则该使用者的手势有效，进行步骤10，否则继续等待检测；
步骤10、将进行上述步骤2到步骤4预处理后的数据再进行步骤6与步骤7处理后，送入神经网络B，判断使用者做了哪种手势，从而进行后续对应的操作，若所做手势不在预先定义的N种手势中，则不产生操作。
以下给出一种在手机终端开启App的应用场景，但不限于此：
手机内置有雷达传感器，并预先存储设备所有者身份信息及手势信息，部分手势信息对应App操作指令；在手机锁屏的前提下，若做手势进行身份验证的人员为设备所有者，则手机解锁，判断手势信息，打开手势对应的App，若无对应手势则不进行任何操作；若手势不属于设备所有者，则手机维持锁屏。在手机已经解锁的情况下，若手势不属于设备所有者或设备所有者所做手势不在规定手势中，也不会进行对应的操作。身份验证和手势识别通过一个手势即可同时作用。这样就完成了整个基于雷达信号的身份及手势识别流程。经过多次实验，证明本发明可以达到90％左右的识别准确率。
综上所述，本发明提供的基于雷达信号的身份及手势识别方法，通过雷达进行信号采集，有效减少能耗，保障用户的隐私。经过一系列有效的预处理后采用神经网络提取特征，能够 根据手势是否属于相应用户，判断是否进行手势对应的操作，这一手段提升了手势识别的安全性。整个流程无需额外的解锁步骤，仅需要一次手势操作，也使得手势识别变得更为便捷。
以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明，不能认定本发明具体实施只局限于这些说明。对于本发明所属技术领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干简单推演或替换，都应当视为属于发明的保护范围。

Claims (8)

  1. 一种基于雷达信号的身份及手势识别方法，其特征在于，该方法包括以下步骤：
    步骤1、读取雷达传感器接收到的经由测试人员各种手势反射得到的雷达回波信号；
    步骤2、将雷达回波信号与雷达发射信号进行混频；
    步骤3、采用高通滤波器对混频后的信号进行滤波；
    步骤4、将滤波后的信号数据进行中心化操作；
    步骤5、设计适用于雷达信号特征的神经网络模型，利用步骤4预处理后的信号与测试人员的身份信息标签对模型进行训练，得到能够进行身份验证的神经网络A；
    步骤6、根据步骤4预处理后的信号及雷达传感器参数，得到距离、速度及角度信息，从而计算得到运动目标的三维坐标及Doppler值；
    步骤7、采用空间栅格法对雷达上方空间进行划分；将空间区域映射为三维矩阵，将步骤6得到的三维坐标及Doppler值根据其所属空间中的栅格位置，将其Doppler值累加在该栅格对应矩阵单元中作为矩阵单元元素的大小；将该矩阵与手势信息标签送入设计好的神经网络中训练得到能够进行手势识别的神经网络B；
    步骤8、将实时检测到的雷达回波信号进行恒虚警检测判断是否有人在进行手势识别，如果有则进行步骤9；无则继续等待检测；
    步骤9、将检测到的雷达回波信号进行步骤2到步骤4的预处理；将处理后的信号送入步骤5中训练好的神经网络A提取特征进行分析；通过特征中含有的身份信息验证使用者的身份，若身份验证通过则该使用者的手势有效，进行步骤10，否则继续等待检测；
    步骤10、将进行步骤2到步骤4预处理后的数据再进行步骤6与步骤7处理后，输入神经网络B判断使用者手势，根据使用者手势与操作的对应关系进行相应操作。
  2. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，所述雷达传感器采用频率区间为57.4GHz至62.6GHz的FMCW毫米波雷达。
  3. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，所述步骤2中，若雷达传感器具有多个接收通道，则将得到的多条对应波形取平均之后再进行滤波。
  4. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，所述步骤3中，采用截止频率为31250Hz的八阶巴特沃斯高通滤波器对混频后的信号进行滤波，在保留有效信息的同时滤除较大的直流和低频噪声。
  5. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，所述 步骤6中，根据步骤4预处理后的信号及雷达传感器参数，通过快时间维度FFT得到距离信息R：
    Figure PCTCN2020083989-appb-100001
    Figure PCTCN2020083989-appb-100002
    再通过慢时间维度FFT得到速度信息v：
    Figure PCTCN2020083989-appb-100003
    Figure PCTCN2020083989-appb-100004
    其中f movingBeat和f staticBeat分别为目标运动和静止状态下差拍信号的频率，f d为多普勒频率，f c为扫频带宽，R为目标距离，C为光速，t c为扫频周期，f为Chirp信号的中心频率，v为目标速度；
    根据雷达传感器的多个收发天线得到角度信息θ：
    Figure PCTCN2020083989-appb-100005
    Figure PCTCN2020083989-appb-100006
    其中Δd为任意两个接收天线距离目标的距离差值，ΔΦ为对应两个天线接收信号的相位差，L是对应两个天线之间的距离，λ为波长。
  6. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，所述步骤7中，将空间大小为Xcm*Ycm*Zcm的区域，映射为一个大小为(X*k1)*(Y*k2)*(Z*k3)的三维矩阵，其中X，Y，Z为空间对应的长宽高，k1，k2，k3为实际空间长宽高与三维矩阵的映射系数。
  7. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，有效使用者在训练时建立规定手势与操作的对应关系，在步骤10识别出有效使用者的具体手势为某一规定手势时，执行该规定手势对应的操作，若识别出的手势不在规定手势中，则不产生操作。
  8. 根据权利要求1所述的一种基于雷达信号的身份及手势识别方法，其特征在于，该方法应用于手机中，手机内置有雷达传感器，并预先存储设备所有者身份信息及手势信息，部分手势信息对应App操作指令；在手机锁屏的前提下，若做手势进行身份验证的人员为设备所有者，则手机解锁，判断手势信息，打开手势对应App，若无对应手势则不进行任何操作；若手势不属于设备所有者，则手机维持锁屏；在手机已经解锁的情况下，若手势不属于设备 所有者或设备所有者所做手势不在规定手势中，则不进行对应的操作。
PCT/CN2020/083989 2020-04-09 2020-04-09 一种基于雷达信号的身份及手势识别方法 WO2021068470A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2020/083989 WO2021068470A1 (zh) 2020-04-09 2020-04-09 一种基于雷达信号的身份及手势识别方法
US17/218,174 US11947002B2 (en) 2020-04-09 2021-03-31 Method for recognizing identity and gesture based on radar signals

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2020/083989 WO2021068470A1 (zh) 2020-04-09 2020-04-09 一种基于雷达信号的身份及手势识别方法

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/218,174 Continuation US11947002B2 (en) 2020-04-09 2021-03-31 Method for recognizing identity and gesture based on radar signals

Publications (1)

Publication Number Publication Date
WO2021068470A1 true WO2021068470A1 (zh) 2021-04-15

Family

ID=75436793

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/083989 WO2021068470A1 (zh) 2020-04-09 2020-04-09 一种基于雷达信号的身份及手势识别方法

Country Status (2)

Country Link
US (1) US11947002B2 (zh)
WO (1) WO2021068470A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI756122B (zh) * 2021-04-30 2022-02-21 開酷科技股份有限公司 距離都卜勒雷達角度感測方法及裝置
CN114563784A (zh) * 2022-04-27 2022-05-31 武汉理工大学 一种基于双毫米波雷达的船载环境入侵检测方法及系统
US11474232B2 (en) 2021-03-19 2022-10-18 KaiKuTek Inc. Range doppler angle detection method and range doppler angle detection device
CN115345908A (zh) * 2022-10-18 2022-11-15 四川启睿克科技有限公司 一种基于毫米波雷达的人体姿态识别方法
CN115497596A (zh) * 2022-11-18 2022-12-20 深圳聚邦云天科技有限公司 一种基于物联网的人体运动过程姿态纠正方法及系统

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113963441B (zh) * 2021-10-25 2024-04-02 中国科学技术大学 一种基于跨域增强的毫米波雷达手势识别方法及系统
CN114511873B (zh) * 2021-12-16 2024-04-23 清华大学 基于毫米波雷达成像的静态手势识别方法及装置
US20230288536A1 (en) * 2022-03-14 2023-09-14 Plume Design, Inc. Radar-based movement sensors arranged in a domestic setting
CN114895275A (zh) * 2022-05-20 2022-08-12 中国人民解放军国防科技大学 基于高效多维注意力神经网络的雷达微动手势识别方法
CN117008075B (zh) * 2023-10-07 2024-01-23 西安乾景防务技术有限公司 一种基于微波天线阵列的无线手势采集方法及系统

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160259037A1 (en) * 2015-03-03 2016-09-08 Nvidia Corporation Radar based user interface
CN106446828A (zh) * 2016-09-22 2017-02-22 西北工业大学 一种基于Wi‑Fi信号的用户身份识别方法
CN106537173A (zh) * 2014-08-07 2017-03-22 谷歌公司 基于雷达的姿势识别
CN107024685A (zh) * 2017-04-10 2017-08-08 北京航空航天大学 一种基于距离‑速度特征的手势识别方法
CN108344995A (zh) * 2018-01-25 2018-07-31 宁波隔空智能科技有限公司 一种基于微波雷达技术的手势识别装置及手势识别方法
CN108469611A (zh) * 2018-03-29 2018-08-31 电子科技大学 基于身份识别的fmcw雷达测距系统基带信号处理方法
CN109188414A (zh) * 2018-09-12 2019-01-11 北京工业大学 一种基于毫米波雷达的手势运动探测方法
CN109271838A (zh) * 2018-07-19 2019-01-25 重庆邮电大学 一种基于fmcw雷达的三参数特征融合手势识别方法
CN110456320A (zh) * 2019-07-29 2019-11-15 浙江大学 一种基于自由空间步态时序特征的超宽带雷达身份识别方法
CN110765974A (zh) * 2019-10-31 2020-02-07 复旦大学 基于毫米波雷达和卷积神经网络的微动手势识别方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8884617B2 (en) * 2008-06-23 2014-11-11 The Regents Of The University Of California Magnetic particle imaging devices and methods
WO2020008458A2 (en) * 2018-07-02 2020-01-09 Vayyar Imaging Ltd. System and methods for environment mapping
WO2020236988A1 (en) * 2019-05-22 2020-11-26 Tactual Labs Co. Millimeter wave array
WO2020263735A1 (en) * 2019-06-25 2020-12-30 OPSYS Tech Ltd. Adaptive multiple-pulse lidar system
KR102305328B1 (ko) * 2019-12-24 2021-09-28 한국도로공사 카메라 기반의 자동화된 정밀도로지도 생성 시스템 및 방법
CN113138557B (zh) * 2020-01-17 2023-11-28 北京小米移动软件有限公司 家居设备控制方法、装置以及存储介质

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106537173A (zh) * 2014-08-07 2017-03-22 谷歌公司 基于雷达的姿势识别
US20160259037A1 (en) * 2015-03-03 2016-09-08 Nvidia Corporation Radar based user interface
CN106446828A (zh) * 2016-09-22 2017-02-22 西北工业大学 一种基于Wi‑Fi信号的用户身份识别方法
CN107024685A (zh) * 2017-04-10 2017-08-08 北京航空航天大学 一种基于距离‑速度特征的手势识别方法
CN108344995A (zh) * 2018-01-25 2018-07-31 宁波隔空智能科技有限公司 一种基于微波雷达技术的手势识别装置及手势识别方法
CN108469611A (zh) * 2018-03-29 2018-08-31 电子科技大学 基于身份识别的fmcw雷达测距系统基带信号处理方法
CN109271838A (zh) * 2018-07-19 2019-01-25 重庆邮电大学 一种基于fmcw雷达的三参数特征融合手势识别方法
CN109188414A (zh) * 2018-09-12 2019-01-11 北京工业大学 一种基于毫米波雷达的手势运动探测方法
CN110456320A (zh) * 2019-07-29 2019-11-15 浙江大学 一种基于自由空间步态时序特征的超宽带雷达身份识别方法
CN110765974A (zh) * 2019-10-31 2020-02-07 复旦大学 基于毫米波雷达和卷积神经网络的微动手势识别方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WANG, YONG, WANG SHA-SHA, TIAN ZENG-SHAN, ZHOU MU, WU JIN-JUN: "Two-Stream Fusion Neural Network Approach for Hand Gesture Recognition Based on FMCW Radar", ACTA ELECTRONICA SINICA, vol. 47, no. 7, 31 July 2019 (2019-07-31), pages 1408 - 1415, XP055798922 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11474232B2 (en) 2021-03-19 2022-10-18 KaiKuTek Inc. Range doppler angle detection method and range doppler angle detection device
TWI756122B (zh) * 2021-04-30 2022-02-21 開酷科技股份有限公司 距離都卜勒雷達角度感測方法及裝置
CN114563784A (zh) * 2022-04-27 2022-05-31 武汉理工大学 一种基于双毫米波雷达的船载环境入侵检测方法及系统
CN114563784B (zh) * 2022-04-27 2022-08-09 武汉理工大学 一种基于双毫米波雷达的船载环境入侵检测方法及系统
CN115345908A (zh) * 2022-10-18 2022-11-15 四川启睿克科技有限公司 一种基于毫米波雷达的人体姿态识别方法
CN115497596A (zh) * 2022-11-18 2022-12-20 深圳聚邦云天科技有限公司 一种基于物联网的人体运动过程姿态纠正方法及系统

Also Published As

Publication number Publication date
US11947002B2 (en) 2024-04-02
US20210318427A1 (en) 2021-10-14

Similar Documents

Publication Publication Date Title
WO2021068470A1 (zh) 一种基于雷达信号的身份及手势识别方法
CN111427031B (zh) 一种基于雷达信号的身份及手势识别方法
Yang et al. MU-ID: Multi-user identification through gaits using millimeter wave radios
CN101551863B (zh) 基于非下采样轮廓波变换的遥感图像道路提取方法
CN108510734A (zh) 一种路侧单元及一种路侧单元的车辆信息匹配方法
CN101404086B (zh) 基于视频的目标跟踪方法及装置
CN109188414A (zh) 一种基于毫米波雷达的手势运动探测方法
CN109101871A (zh) 一种基于深度和近红外信息的活体检测装置、检测方法及其应用
CN106646457A (zh) 基于微波的人员行为检测方法与装置
CN102945374B (zh) 一种高分辨率遥感图像中民航飞机自动检测方法
CN113408328B (zh) 基于毫米波雷达的手势分割与识别算法
CN108446696A (zh) 一种基于深度学习的端到端车牌识别方法
CN108306842A (zh) 售车过程中无人陪同试驾的系统及方法
CN103177458A (zh) 一种基于频域分析的可见光遥感图像感兴趣区域检测方法
WO2022116467A1 (zh) 一种基于毫米波雷达的非接触式身份识别方法及系统
Chen et al. Through-wall human motion recognition based on transfer learning and ensemble learning
CN105427342A (zh) 一种水下小目标声呐图像目标检测跟踪方法和系统
CN101533466B (zh) 一种眼睛定位图像处理方法
CN108307767A (zh) 适用于全自动除草机的障碍物检测避障系统及方法
CN109919960A (zh) 一种基于多尺度Gabor滤波器的图像连续边缘检测方法
CN113391282A (zh) 一种基于雷达多维特征融合的人体姿态识别方法
Chen et al. Multi-view real-time human motion recognition based on ensemble learning
Rizik et al. Cost-efficient FMCW radar for multi-target classification in security gate monitoring
CN113837131A (zh) 一种基于fmcw毫米波雷达的多尺度特征融合手势识别方法
CN108932468A (zh) 一种适用于心理学的面部识别方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20875018

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20875018

Country of ref document: EP

Kind code of ref document: A1