CN110705501B - Interference suppression method for improving gesture recognition precision of FMCW radar - Google Patents
Interference suppression method for improving gesture recognition precision of FMCW radar Download PDFInfo
- Publication number
- CN110705501B CN110705501B CN201910971230.5A CN201910971230A CN110705501B CN 110705501 B CN110705501 B CN 110705501B CN 201910971230 A CN201910971230 A CN 201910971230A CN 110705501 B CN110705501 B CN 110705501B
- Authority
- CN
- China
- Prior art keywords
- frame
- doppler
- matrix
- rdm
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- 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
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
- G01S13/583—Velocity or trajectory determination systems; Sense-of-movement determination systems using transmission of continuous unmodulated waves, amplitude-, frequency-, or phase-modulated waves and based upon the Doppler effect resulting from movement of targets
-
- 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
- G01S7/35—Details of non-pulse systems
- G01S7/352—Receivers
- G01S7/354—Extracting wanted echo-signals
Abstract
The invention provides an interference suppression method for improving gesture recognition accuracy of an FMCW radar. Firstly, collecting frequency sweep data of each gesture action by using an FMCW radar, carrying out FFT (fast Fourier transform) on the radar data of each frequency sweep to obtain distance estimation of a gesture target, and carrying out FFT on the radar frequency sweep data of each chirp on the basis of the previous FFT to obtain Doppler estimation of the gesture target; secondly, coupling the distance estimation and the Doppler estimation to obtain the RDM of the gesture target, and then removing the background noise of the RDM by using a frame difference method; then static and dynamic interference suppression is carried out; and finally, inputting the RDM image subjected to interference suppression into a depth 3-dimensional convolution (3D ConvNet, C3D) network, an expanded 3D convolution (I3D ConvNet, I3D) network and a time sequence expanded 3D convolution ((Long Short Term Memory network-expanded 3D ConvNet, TS-I3D) network, carrying out feature extraction and classification, and outputting different gesture types.
Description
Technical Field
The invention relates to the field of human-computer interaction, in particular to an interference suppression method for improving gesture recognition accuracy of an FMCW radar.
Background
With the development of man-machine interaction technology in the modern times, gesture recognition becomes an important component of man-machine interaction, the research and development of the gesture recognition influence the naturalness and flexibility of the man-machine interaction, and the gesture recognition is widely applied to various fields. In the aspect of home entertainment, characters in a game are controlled according to actions of a user such as swinging left and right in a game environment, so that the user experience effect is better. In the aspect of intelligent driving, the control over the navigation system and the vehicle-mounted entertainment system is completed by recognizing the gesture actions of the driver, so that the driving safety can be improved.
Because the traditional gesture recognition method mainly utilizes an optical camera and a depth camera, the influence of abnormal illumination cannot be overcome, and the privacy of a user cannot be effectively protected, the FMCW radar has a very considerable research prospect in the application of gesture recognition.
In the current research of gesture recognition based on radar, only the distance and Doppler information of a gesture target are mined, but noise and interference existing in the motion process of the gesture target are ignored, a lot of noise and interference exist in measured radar data, static target interference mainly comprises a wall body, a table and a chair, obstacles and the like, and dynamic target interference mainly comprises pedestrians, user arms, bodies and the like. Therefore, the interference suppression is a very influential key step in the gesture recognition, and can effectively improve the accuracy of the gesture recognition.
Disclosure of Invention
The invention aims to provide an interference suppression method for improving gesture recognition accuracy of an FMCW radar. Compared with the traditional interference suppression method, the method combines background noise, effectively suppresses static interference and dynamic interference, makes the motion information of the gesture actions after suppression more obvious, and improves the accuracy of gesture recognition.
The technical scheme adopted by the invention is as follows: an interference suppression method for improving gesture recognition accuracy of an FMCW radar specifically comprises the following steps:
step one, setting Frequency Modulated Continuous Wave (FMCW) radar parameters, configuring an antenna to transmit 2 and receive 4, wherein the number of Frequency sweeps is N c Number of sampling points is N s The frame number is frame. Then, collecting data of each gesture action by using a radar, wherein each frame of data is stored as N S ×N c Matrix F of k ;
Step two, accumulating the data of 2 transmitting antennas and 4 receiving antennas, obviously enhancing the signal-to-noise ratio of the signal, and then carrying out F k Performs Fourier transform on each column of data to generate a distance spectrum matrix P r ;
Step three, distance spectrum matrix P r Performs Fourier transform on each row of data to generate a Doppler matrix S k ;
Step four, distance spectrum matrix P of each frame r And S k Coupling to obtain a Range-Doppler Map (RDM) of each frame of gesture target;
step five, accumulating a plurality of frame data and obtaining an average value as a background frame X bk In the present invention, the number of frames taken is 3. Subtracting the background frame by frame difference method to obtainWherein, the elements of the ith row and the jth column in the RDM matrix to be processed areBackground frame X bk The ith row and the jth column ofm and n represent the total number of pixels of the RDM from the distance axis and the Doppler axis respectively;
step six, removing noise in RDM by adopting a self-adaptive threshold, comparing elements in the RDM of the detected unit with the threshold, and if the element value in the RDM of the detected unit is larger than a threshold value N v If the target is detected, judging that the target is detected, and setting a processing result matrix as R;
seventhly, static interference suppression is carried out on the matrix R, and the result matrix is recorded as R sta ;
Step eight, pairing the matrix R sta Carrying out dynamic interference suppression, marking a gesture target, and recording a result matrix as R res ;
Step nine, sending the 32-frame RDM into a C3D, I3D and TS-I3D network for feature extraction to obtain a feature function F fusion And F is fusion And inputting a softmax classifier for classification, and outputting different gesture categories.
The fifth step comprises the following steps:
5.1 in each gesture motion data sample, accumulating the s frame data and obtaining the average value as the background frame, and expressing as:
wherein s represents the accumulated number of RDM background frames, and m and n are the total number of pixels of the Doppler axis and the distance axis of the RDM respectively;
5.2 decompose the RDM background frame into two parts:
wherein, the first term represents the area where the doppler frequency offset is not zero, and the 2 nd term represents the area where the doppler frequency offset is zero. r and d represent the abscissa and ordinate of the RDM, characterizing the distance and doppler axes, respectively;
5.3 because the interference of the static target is introduced into the background frame, in order to avoid the influence of the interference of the static target on the calculation of the background frame, the calculation mode of the background frame is changed into:
5.4 considering the fluctuation of the background noise, which results in that the noise cannot be removed uniformly in each frame subtraction process, the calculation method of the background frame subtraction operation is defined as:
wherein x is p Representing the pixel value, x, of each frame pk Representing the pixel values of the background frame.
In the sixth step, a self-adaptive threshold is adopted to remove noise, and the noise threshold is as follows:
wherein the content of the first and second substances,and elements of an r-th row and a d-th column in the RDM to be processed are shown. Comparing the detected elements in the RDM with a threshold, and if the value of the elements in the RDM of the detected unit is larger than the threshold value N v If yes, the target is judged, and the result matrix is marked as R.
The seventh step comprises the following steps:
7.1 statistics of the number of targets detected in R, denoted as c d ;
7.2 search for the target with the Doppler frequency offset value of zero in R, and the number is recorded as c s Distance coordinate vector is noted as B cur . Similarly, recording the distance unit vector B of the target in the previous frame pre ;
7.3 pairs of B cur And B pre Sorting in ascending order, if c d And c s Count unequal, find all B cur And B pre The same coordinate distance is obtained, and the targets with the same coordinate distance are deleted from the R to obtain a matrix R sta 。
The eighth step comprises the following steps:
8.1 recording R sta The coordinates of the target. C x Recording the distance coordinates x of all targets i ,C y Recording the Doppler coordinate y of all targets i ,C loc Record R sta The distance value and Doppler value of the ith target are (x) i ,y i );
8.2 pairs of C x All elements in the list are sorted in ascending order, and C is added x The minimum value is recorded as m x (ii) a To C y Counting all Doppler coordinates, and recording the number of the forward Doppler speeds as c p The number of negative Doppler velocities is c n ;
8.3 definition of vector M y For recording doppler values that satisfy the conditions. Find C loc Each coordinate (x) of i ,y i ) If x is i =m x At M y Middle record Doppler value y i If c, then not at M y Middle record Doppler value y i ;
8.4 definition of m y Is a doppler frequency offset bin value. If c is p >c n Let m stand for y Is equal to M y Maximum value of, otherwise let m y Is equal to M y Minimum value of (1);
8.5 coordinate is (m) x ,m y ) The element(s) of (1) is (are) marked as a gesture target, and the result matrix is (are) marked as R res 。
Drawings
FIG. 1 is a flow chart of gesture recognition according to the present invention;
FIG. 2 is a schematic diagram of the RDM construction of the present invention;
FIG. 3 is a diagram of an example of a continuous 16-frame RDM according to the present invention;
FIG. 4 is a front and back comparison of a background frame calculation method according to the present invention;
FIG. 5 is a diagram of the background noise removal result of a continuous 3-frame RDM according to the present invention;
FIG. 6 is a diagram of the static and dynamic interference suppression effect of a single frame RDM according to the present invention;
FIG. 7 is a diagram of the accumulation of successive 16 RDMs before and after interference suppression according to the present invention;
FIG. 8 illustrates the gesture recognition accuracy of the present invention in different networks;
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
step one, setting Frequency Modulated Continuous Wave (FMCW) radar parameters, configuring an antenna to transmit 2 and receive 4, wherein the number of Frequency sweeps is N c Number of sampling points is N s The frame number is frame. Then, collecting data of each gesture action by using a radar, wherein each frame of data is stored as N S ×N c Matrix F of k ;
The transmitted signal of FMCW is:
wherein the content of the first and second substances,is a linear function of the frequency of the transmitted signal over time, f c Is the carrier frequency, B is the bandwidth, A TX Is the amplitude of the transmitted signal and T is the period. Obtaining echo signals through time delay and gesture motion errors, mixing the echo signals with the sending signals, and obtaining intermediate-frequency signals s through a low-pass filter IF (t):
The distance estimate may be calculated from the intermediate frequency signal, where f IF Representing the intermediate frequency signal frequency, c is the speed of light, B is the bandwidth, T is the period:
step two, by coherent accumulation, for F k Performs Fourier transform on each column of data to generate a distance spectrum matrix P r ;
Step three, distance spectrum matrix P r Performing Fourier transform on each row of data, performing Doppler estimation by using phase change of multiple repeated sweep data, and generating a Doppler matrix S k ;
And step four, accumulating the frequency sweep to form two dimensions of a slow time domain and a fast time domain. Fast time domain corresponding distance spectrum matrix P r Slow time domain corresponding Doppler matrix S k Obtaining the RDM of the gesture target after coupling;
step five, accumulating a plurality of frame data and obtaining an average value as a background frame, and subtracting the background frame by using a frame difference method to weaken the noise of the RDM, wherein the method specifically comprises the following steps:
5.1 in each gesture motion data sample, accumulating a plurality of frame data and obtaining an average value as a background frame, wherein the background frame is expressed as:
where s represents the cumulative number of RDM background frames, and m and n are the total number of pixels from the doppler axis and the distance axis of the RDM, respectively.
5.2 decompose the RDM background frame into two parts:
wherein, the first term represents the area of non-zero Doppler frequency offset, and the 2 nd term represents the area of zero Doppler frequency offset. r and d represent the row and column coordinates of the RDM, respectively characterizing the range and doppler axes;
5.3 because the interference of the static target is introduced into the background frame, in order to avoid the influence of the interference of the static target on the calculation of the background frame, the calculation mode of the background frame is changed into:
5.4 considering the fluctuation of the background noise, which results in that the noise cannot be removed uniformly in each frame subtraction process, the calculation method of the background frame subtraction operation is defined as:
step six, after the background frame is utilized to weaken the noise, the noise in the RDM is removed in a self-adaptive mode by adopting a self-adaptive threshold, and the noise threshold is as follows:
wherein the content of the first and second substances,and elements of an r-th row and a d-th column in the RDM to be processed are shown. Comparing the level value of the detected unit with a threshold, and if the element value in the RDM of the detected unit is larger than the threshold value N v If yes, the target is judged to be detected, and the processing result matrix is R.
Step seven, performing static interference suppression on the processing result obtained in the step six, and specifically comprising the following steps:
7.1 inhibition statistics the number of targets detected in R, denoted c d ;
7.2 search for the target with Doppler frequency offset value of zero in R, and number is recorded as c s Distance coordinate vector is denoted as B cur . Similarly, recording the distance unit vector B of the target in the previous frame pre ;
7.3 pairs of B cur And B pre Sorting in ascending order, if c d And c s Count unequal, find all B cur And B pre The same coordinate distance is obtained, and the targets with the same coordinate distance are deleted from the R to obtain a matrix R sta 。
Step eight, performing dynamic interference suppression on the processing result obtained in the step seven, and specifically comprising the following steps:
8.1 recording R sta The coordinates of the target. C x Recording the distance coordinates x of all objects i ,C y Recording the Doppler coordinate y of all targets i ,C loc Record R sta The distance value and Doppler value of the ith target are (x) i ,y i );
8.2 pairs of C x All elements in the list are sorted in ascending order, and C is added x The minimum value is recorded as m x (ii) a To C y Counting all Doppler coordinates, and recording the number of the forward Doppler speeds as c p The number of negative Doppler velocities is c n ;
8.3 definition of vector M y For recording doppler values that satisfy the conditions. Find C loc Each coordinate (x) of i ,y i ) If x is i =m x At M y Middle record Doppler value y i If x i ≠m x Then is not in M y Middle record Doppler value y i ;
8.4 definition of m y Is a doppler frequency offset bin value. If c is p >c n Let m stand for y Is equal to M y Maximum value of, otherwise let m y Is equal to M y Minimum value of (d);
8.5 coordinate is (m) x ,m y ) The element(s) of (1) is (are) marked as a gesture target, and the result matrix is (are) marked as R res 。
Claims (4)
1. An interference suppression method for improving gesture recognition accuracy of an FMCW radar is characterized by comprising the following steps of:
step one, setting frequency modulation continuous wave radar parameters, configuring an antenna to send and receive 2, wherein the frequency sweeping number is N c Number of sampling points is N s The frame number is frame; then, collecting data of each gesture action by using a radar, wherein each frame of data is stored as N S ×N c Matrix F of k ;
Step two, accumulating the data of 2 transmitting antennas and 4 receiving antennas, which can obviously enhance the signal-to-noise ratio of the signal, and then for F k Performs Fourier transform on each column of data to generate a distance spectrum matrix P r ;
Step three, distance spectrum matrix P r Performs Fourier transform on each row of data to generate a Doppler matrix S k ;
Step four, distance spectrum matrix P of each frame r And S k Coupling to obtain a distance-Doppler matrix of each frame of gesture target, namely an RDM matrix;
step five, accumulating a plurality of frame data and obtaining an average value as a background frame X bk Taking the frame number as 3, and subtracting the background frame by using a frame difference method to obtain the frame differenceWherein, the elements of the ith row and the jth column in the RDM matrix to be processed areBackground frame X bk The ith row and the jth column ofm and n respectively represent the total number of pixels on the Doppler axis and the distance axis of the RDM matrix;
removing noise in the RDM matrix by adopting a self-adaptive threshold, comparing elements in the RDM matrix of the detected unit with the threshold, if the element value in the RDM matrix of the detected unit is greater than the threshold value, judging that a target is detected, and processing the result matrix to be R;
seventhly, static interference suppression is carried out on the matrix R, and the result matrix is recorded as R sta ;
Step eight, the matrix R is paired sta Carrying out dynamic interference suppression, marking a gesture target, and recording a result matrix as R res ;
Step nine, sending the 32-frame RDM into a C3D, I3D and TS-I3D network for feature extraction to obtain a feature function F fusion And F is fusion Inputting a softmatx classifier for classification, and outputting different gesture categories;
the fifth step comprises the following steps:
5.1 in each gesture motion data sample, accumulating the s frame data and obtaining the average value as the background frame, and expressing as:
wherein s represents the cumulative number of RDM background frames, and m and n are the total number of pixels of the distance axis and Doppler axis of RDM respectively;
5.2 decompose the RDM background frame into two parts:
the first term represents a region with non-zero Doppler frequency offset, the 2 nd term represents a region with zero Doppler frequency offset, and r and d represent horizontal and vertical coordinates of RDM (remote data management) and represent a distance and a Doppler axis respectively;
5.3 because the interference of the static target is introduced into the background frame, in order to avoid the influence of the interference of the static target on the calculation of the background frame, the calculation mode of the background frame is changed into:
5.4 considering the fluctuation of the background noise, which results in that the noise cannot be removed uniformly in each frame subtraction process, the calculation method of the background frame subtraction operation is defined as:
wherein x is p Representing the pixel value, x, of each frame bk Representing the pixel values of the background frame.
2. The interference suppression method for improving the gesture recognition accuracy of the FMCW radar as claimed in claim 1, wherein the noise is removed by using an adaptive threshold in step six, where the noise threshold is:
wherein, the first and the second end of the pipe are connected with each other,representing the elements of the r-th row and the d-th column in the RDM to be processed, if the element value in the RDM of the detected unit is larger than the threshold value N v If yes, the target is judged, and the result matrix is marked as R.
3. The interference suppression method for improving gesture recognition accuracy of frequency modulated continuous wave radar according to claim 1, wherein the seventh step comprises the following steps:
7.1 statistics of the number of targets detected in R, denoted as c d ;
7.2 search for the target with the Doppler frequency offset value of zero in R, and the number is recorded as c s Distance coordinate vector is noted as B cur Recording the distance unit vector B of the target in the previous frame pre ;
7.3 pairs of B cur And B pre Respectively sorting in ascending order if c d And c s Count inequality, find all B cur And B pre The same coordinate distance is obtained, and the targets with the same coordinate distance are deleted from the R to obtain a matrix R sta 。
4. The interference suppression method for improving gesture recognition accuracy of frequency modulated continuous wave radar according to claim 1, wherein the eighth step comprises the following steps:
8.1 recording R sta Coordinates of the target; c x Recording the distance coordinates x of all targets i ,C y Recording the Doppler coordinate y of all targets i ,C loc Record R sta The distance value and Doppler value of the ith target are (x) i ,y i );
8.2 pairs of C x All elements in the list are sorted in ascending order, and C is added x The minimum value is recorded as m x (ii) a To C y Counting all Doppler coordinates in the sample, and recording the number of the forward Doppler speeds as c p The number of negative Doppler velocities is c n ;
8.3 definition of vector M y Used for recording Doppler value meeting the condition; find C loc Each coordinate (x) of i ,y i ) If x is i =m x At M y Middle record Doppler value y i If x is i ≠m x Then is not in M y Middle record Doppler value y i ;
8.4 definitionm y Is a Doppler frequency offset cell value; if c is p >c n Let m stand for y Is equal to M y Otherwise let m be y Is equal to M y Minimum value of (1);
8.5 coordinates are (m) x ,m y ) The element(s) of (1) is (are) marked as a gesture target, and the result matrix is (are) marked as R res 。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910971230.5A CN110705501B (en) | 2019-10-14 | 2019-10-14 | Interference suppression method for improving gesture recognition precision of FMCW radar |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910971230.5A CN110705501B (en) | 2019-10-14 | 2019-10-14 | Interference suppression method for improving gesture recognition precision of FMCW radar |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110705501A CN110705501A (en) | 2020-01-17 |
CN110705501B true CN110705501B (en) | 2022-09-20 |
Family
ID=69198829
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910971230.5A Active CN110705501B (en) | 2019-10-14 | 2019-10-14 | Interference suppression method for improving gesture recognition precision of FMCW radar |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110705501B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11639985B2 (en) | 2020-07-02 | 2023-05-02 | International Business Machines Corporation | Three-dimensional feature extraction from frequency modulated continuous wave radar signals |
CN111813222B (en) * | 2020-07-09 | 2022-03-29 | 电子科技大学 | Terahertz radar-based fine dynamic gesture recognition method |
CN112285656B (en) * | 2020-09-27 | 2024-02-09 | 西安空间无线电技术研究所 | Time-sensitive target deception jamming prevention method based on track characteristics |
CN112764002B (en) * | 2021-01-07 | 2022-09-30 | 北京理工大学重庆创新中心 | FMCW radar gesture recognition method based on deformable convolution |
CN113064367B (en) * | 2021-03-16 | 2022-01-04 | 矽典微电子(上海)有限公司 | Toilet bowl control method and system based on millimeter wave radar |
CN113313040B (en) * | 2021-06-04 | 2022-05-13 | 福州大学 | Human body posture identification method based on FMCW radar signal |
CN113591684A (en) * | 2021-07-29 | 2021-11-02 | 北京富奥星电子技术有限公司 | Gesture recognition method based on Doppler radar of CW system |
CN113687348B (en) * | 2021-09-18 | 2024-03-15 | 无锡威孚高科技集团股份有限公司 | Pedestrian recognition method and device based on tracking micro Doppler graph |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106022227A (en) * | 2016-05-11 | 2016-10-12 | 苏州大学 | Gesture identification method and apparatus |
CN106970647A (en) * | 2017-05-27 | 2017-07-21 | 吉林大学 | Unmanned vehicle gesture interaction device based on RDM algorithms |
CN108509910A (en) * | 2018-04-02 | 2018-09-07 | 重庆邮电大学 | Deep learning gesture identification method based on fmcw radar signal |
CN109188414A (en) * | 2018-09-12 | 2019-01-11 | 北京工业大学 | A kind of gesture motion detection method based on millimetre-wave radar |
CN109271838A (en) * | 2018-07-19 | 2019-01-25 | 重庆邮电大学 | A kind of three parameter attributes fusion gesture identification method based on fmcw radar |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10481696B2 (en) * | 2015-03-03 | 2019-11-19 | Nvidia Corporation | Radar based user interface |
US20190049558A1 (en) * | 2017-08-08 | 2019-02-14 | KC Sense Limited | Hand Gesture Recognition System and Method |
-
2019
- 2019-10-14 CN CN201910971230.5A patent/CN110705501B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106022227A (en) * | 2016-05-11 | 2016-10-12 | 苏州大学 | Gesture identification method and apparatus |
CN106970647A (en) * | 2017-05-27 | 2017-07-21 | 吉林大学 | Unmanned vehicle gesture interaction device based on RDM algorithms |
CN108509910A (en) * | 2018-04-02 | 2018-09-07 | 重庆邮电大学 | Deep learning gesture identification method based on fmcw radar signal |
CN109271838A (en) * | 2018-07-19 | 2019-01-25 | 重庆邮电大学 | A kind of three parameter attributes fusion gesture identification method based on fmcw radar |
CN109188414A (en) * | 2018-09-12 | 2019-01-11 | 北京工业大学 | A kind of gesture motion detection method based on millimetre-wave radar |
Non-Patent Citations (4)
Title |
---|
24 GHz FMCW Radar System for Real-Time Hand Gesture Recognition Using LSTM;Jun Seuk Suh等;《2018 Asia-Pacific Microwave Conference (APMC)》;20190117;全文 * |
Two Dimensional Parameters Based Hand Gesture Recognition Algorithm for FMCW Radar Systems;Yong Wang等;《International Conference on Wireless and Satellite Systems》;20190507;全文 * |
基于FMCW雷达的双流融合神经网络手势识别方法;王勇等;《电子学报》;20190715;全文 * |
基于FMCW雷达的多维参数手势识别算法;王勇等;《电子与信息学报》;20180917;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN110705501A (en) | 2020-01-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110705501B (en) | Interference suppression method for improving gesture recognition precision of FMCW radar | |
Brodeski et al. | Deep radar detector | |
CN111045008B (en) | Vehicle millimeter wave radar target identification method based on widening calculation | |
US20200256999A1 (en) | Lidar techniques for autonomous vehicles | |
CN109581378B (en) | Moving object detection method, electronic device, and storage medium | |
CN110501685B (en) | Multiframe phase-coherent accumulation method based on radar signal phase compensation | |
CN111323756B (en) | Marine radar target detection method and device based on deep learning | |
CN111323757B (en) | Target detection method and device for marine radar | |
CN115061113B (en) | Target detection model training method and device for radar and storage medium | |
Kalyan et al. | A random finite set based detection and tracking using 3D LIDAR in dynamic environments | |
US20230139751A1 (en) | Clustering in automotive imaging | |
CN108196238B (en) | Clutter map detection method based on adaptive matched filtering under Gaussian background | |
CN110658506B (en) | Micro Doppler clutter filtering method based on angle clustering and Doppler analysis | |
CN109239677B (en) | Environment self-adaptive constant false alarm rate detection threshold determination method | |
CN106603976B (en) | Intelligent microwave frequency band radio monitoring control system | |
CN112835006B (en) | Method and system for detecting small targets on sea by tracking radar based on interframe accumulation | |
Kim et al. | Deep-learning based multi-object detection and tracking using range-angle map in automotive radar systems | |
CN110632592B (en) | False alarm eliminating method for handheld through-wall radar | |
CN112166341B (en) | Speed determination method, apparatus and storage medium | |
CN113885006A (en) | Method for detecting coherent accumulation of maneuvering turning target | |
Stevens et al. | Automated gain control through deep reinforcement learning for downstream radar object detection | |
KR101619064B1 (en) | Target tracking method using active clutter map | |
CN113960587B (en) | Millimeter wave radar multi-target tracking method based on category information feedback | |
CN116299299B (en) | Speed disambiguation method, device, radar equipment and storage medium | |
CN115496170B (en) | Human body posture recognition method and system, electronic equipment and storage medium |
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 |