Radar gesture recognition method, system and device and electronic equipment

Abstract

The invention discloses a radar gesture recognition method, system and device and electronic equipment, and the method comprises the following steps: S1, transmitting a frequency modulation continuous wave signal to a measurement region, receiving an echo signal scattered by the measurement region, and converting the simulated echo signal into a digital signal containing gesture information; s2, performing data processing on the digital signal containing the gesture information to generate original radar point cloud data containing a gesture target; s3, effective radar point cloud data are screened, point cloud features are generated, whether gestures exist or not is judged, if yes, the step S4 is executed, and if not, the step S1 is executed; s4, a rule matching algorithm without machine learning and a template library is used for directly judging which type of gesture features the gesture features accord with, and the gesture type is determined; and S5, displaying output results of feature extraction and gesture discrimination to realize real-time visual interaction. The method has the advantages that machine learning and a template library are not needed, use is easy, and cost is low.

Description

Technical Field

[0001] The present invention belongs to the technical field of human-computer interaction, and in particular relates to a radar gesture recognition method, system, device and electronic equipment. Background Art

[0002] With the rapid development and widespread application of radar technology, radar gesture recognition has become an important technology in the field of human-computer interaction. Compared with traditional optical sensors, radar sensors can work normally in bad weather conditions such as rain, snow, fog or dark conditions, and have the advantages of all-weather and all-day. In addition, radar signals also have great advantages in protecting user privacy information.

[0003] The performance of visual sensors is significantly affected by environmental factors such as lighting, background complexity, and occlusion. In low-light or strong light conditions, the accuracy of gesture recognition may drop significantly. In addition, the complexity of the background can also interfere with the recognition process, resulting in misjudgment or missed judgment.

[0004] Radar point cloud contains rich information such as target spatial position, scattering amplitude, Doppler, etc., which can be used to extract the spatial structure and motion characteristics of gestures, which is conducive to the accurate tracking and recognition of gestures.

[0005] Existing radar gesture recognition methods require machine learning methods to train classification models or match with pre-set template libraries, which requires a lot of data collection and storage, making it difficult to be supported by a variety of edge embedded devices and increasing the cost and complexity of the system. Existing technologies may perform poorly when processing diverse gestures, especially when the speed, angle, and frequency of gestures vary greatly. Traditional methods often have difficulty adapting to new gestures or changing gesture patterns, resulting in reduced recognition rates. Summary of the invention

[0006] In view of the deficiencies in the prior art, the object of the present invention is to provide a radar gesture recognition method, system, device and electronic equipment to solve the above-mentioned technical problems existing in the prior art.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A radar gesture recognition method comprises the following steps:

[0009] S1, transmitting a frequency modulated continuous wave signal to a measurement area, receiving an echo signal scattered from the measurement area, and converting the analog echo signal into a digital signal containing gesture information;

[0010] S2, performing data processing on the digital signal containing the gesture information to generate original radar point cloud data containing the gesture target;

[0011] S3: Filter valid radar point cloud data, generate point cloud features, and determine whether there is a gesture. If yes, proceed to S4, if not, return to S1;

[0012] S4: Use a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the gesture features match and determine the gesture category;

[0013] S5: Display the output results of feature extraction and gesture recognition to achieve real-time visual interaction.

[0014] Furthermore, the S3 specifically includes the following steps:

[0015] S31, determining whether each data point in the original radar point cloud of the current frame is within the gesture interaction area of ​​interest; if so, retaining it; if not, discarding it;

[0016] S32, clustering the three-dimensional spatial position coordinates processed by S31, and determining whether the number of clusters obtained is greater than 0; if so, retaining the cluster with the largest number of data points, discarding other data points, and proceeding to S33; if not, returning to S31;

[0017] S33, calculating the mean position of the three-dimensional coordinates of all data points in the cluster, and calculating the distance from all data points to the mean position, and judging whether it is less than the set threshold; if so, retain it and proceed to S34, if not, discard it;

[0018] S34, calculating the target parameters of the current frame using all the data points filtered by S31, S32 and S33;

[0019] S35, sending the target parameters of the current frame into a sliding window of a certain frame length to form a gesture feature;

[0020] S36. Use some target parameters in the sliding window to determine whether there is a gesture; if so, proceed to S4; if not, return to S1.

[0021] Furthermore, S4 does not need to collect a large number of gesture samples in advance to train the machine learning classification model, and does not need to set a gesture template library in advance. It only needs to use the set gesture judgment rules to determine which type of gesture feature the point cloud features obtained in step S3 meet, thereby determining the gesture category.

[0022] The S4 specifically includes the following steps:

[0023] S41. Determine whether each frame is a gesture frame by using the target parameters of the frame in the sliding window.

[0024] S42, defining the first gesture frame in the sliding window as a starting gesture frame, and calculating the three-dimensional position of the starting gesture frame as the average of the three-dimensional positions of all data points contained in the frame as the starting gesture position;

[0025] S43, calculating the mean of the three-dimensional positions of all data points in all gesture frames except the starting gesture frame as the central gesture position;

[0026] S44. Define the phase center of the two-dimensional antenna array of the radar as the origin of the three-dimensional rectangular coordinate system, define the azimuth, pitch and antenna array direction as the X direction, Z direction and Y direction respectively, calculate the difference between the center gesture position and the starting gesture position in the three coordinate directions, and the direction corresponding to the maximum difference is the gesture movement direction; define the +X direction gesture and the -X direction gesture as the left gesture and the right gesture respectively, define the +Z direction gesture and the -Z direction gesture as the upward gesture and the downward gesture respectively, and define the +Y direction gesture and the -Y direction gesture as the forward push gesture and the backward pull gesture respectively.

[0027] A radar gesture recognition system, comprising:

[0028] A radar subsystem, used for transmitting a frequency modulated continuous wave signal to a measurement area, receiving an echo signal scattered from the measurement area, and converting the analog echo signal into a digital signal that can be processed by a data preprocessing subsystem;

[0029] A data preprocessing subsystem is used to process digital signals and generate raw radar point cloud data containing gesture targets;

[0030] Feature extraction subsystem, used to filter valid radar point cloud data, generate point cloud features, and determine whether there is a gesture;

[0031] The gesture recognition subsystem uses a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the point cloud features match and determine the gesture category;

[0032] The human-computer interaction subsystem is used to display the output results of the feature extraction subsystem and the gesture recognition subsystem, and to realize the real-time visual interaction between human gestures and the radar gesture recognition system.

[0033] A radar gesture recognition device, comprising:

[0034] The radar RF front-end module is placed in front of the observation gesture, transmits a frequency modulated continuous wave signal to the measurement area, receives the echo signal scattered from the measurement area, and converts the analog echo signal into a digital signal that can be processed by the data preprocessing module;

[0035] The data preprocessing module processes the digital signal and generates the original radar point cloud data containing the gesture target;

[0036] Feature extraction module, which filters valid radar point cloud data, generates point cloud features, and determines whether there is a gesture;

[0037] The gesture recognition module uses a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the point cloud features match and determine the gesture category;

[0038] The human-computer interaction module displays the output results of the feature extraction module and the gesture recognition module, and realizes the real-time visual interaction between human gestures and the radar gesture recognition device.

[0039] Furthermore, the radar RF front-end module includes a radar signal transmitting component, a radar signal receiving component, a mixer and an analog-to-digital conversion component;

[0040] The radar signal transmitting component transmits a frequency modulated continuous wave signal to the measurement area, the radar signal receiving component collects the echo signal scattered from the measurement area, and the mixer and the analog-to-digital conversion component convert the analog echo signal into digital echo data.

[0041] Furthermore, the radar RF front-end module includes at least one transmitting antenna and three receiving antennas, and the three receiving antennas are arranged in an L-shaped array, and azimuth and elevation angle measurements are simultaneously achieved, so as to obtain a target point cloud in three-dimensional space;

[0042] Or at least two transmitting antennas and two receiving antennas are included to realize an equivalent two-dimensional antenna array.

[0043] Furthermore, the signal transmitted by the radar RF front-end module is a frequency modulated continuous wave system.

[0044] An electronic device comprises a processing unit, a display unit and a communication unit.

[0045] The processing unit implements the method when executing the executable program;

[0046] The display unit is used to display the intermediate process and the final result;

[0047] The communication unit is used for communication within the system and communication with the outside of the system.

[0048] Furthermore: it also includes a storage unit, the storage medium stores a feasibility program, and the executable program is executed by the processing unit to execute the method described.

[0049] Beneficial effects of the present invention:

[0050] 1. The present invention acquires gesture information by transmitting a frequency modulated continuous wave signal and receiving a scattered echo signal, converts the analog echo signal into a digital signal, and generates original radar point cloud data. Finally, through a multi-step screening process, only valid point cloud data is retained to determine the existence of gestures.

[0051] 2. The radar gesture recognition method adopted by the present invention does not require pre-collection of data to train the machine learning classification model and establish a classification template library. It can directly use the set rule matching algorithm to realize accurate recognition of various gestures, which is more conducive to the use of different scenes and different objects.

[0052] 3. The radar gesture recognition method of the present invention can be implemented using various types of miniaturized embedded devices with lower costs and simpler devices, without the need for complex and expensive processing devices, and is easier to deploy and promote. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0054] Figure 1 Schematic diagram of a radar gesture recognition method according to an embodiment of the present invention;

[0055] Figure 2 This is a flow chart of step 3 of a radar gesture recognition method in an embodiment of the present invention;

[0056] Figure 3 A schematic diagram of six types of gestures designed in an embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram of the structure of a radar gesture recognition system in an embodiment of the present invention;

[0058] Figure 5 This is a schematic diagram of the structure of a radar gesture recognition device in an embodiment of the present invention;

[0059] Figure 6 Schematic diagram of the structure of a radar radio frequency front-end module in a radar gesture recognition device according to an embodiment of the present invention;

[0060] Figure 7 Schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The technical scheme in the embodiment of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiment of the present invention. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] like Figure 1 As shown, an embodiment of the present invention provides a radar gesture recognition method, the method comprising:

[0063] S1: transmitting a frequency modulated continuous wave signal to the measurement area, receiving an echo signal scattered from the measurement area, and converting the analog echo signal into a digital signal containing gesture information;

[0064] S2: Processing the digital signal containing the gesture information to generate original radar point cloud data containing the gesture target;

[0065] S3: Filter valid radar point cloud data, generate point cloud features, and determine whether there is a gesture. If so, proceed to S4; if not, return to S1.

[0066] Optionally, S3 includes the following steps (see Figure 2 ):

[0067] S31: Determine whether each data point in the original radar point cloud of the current frame is within the gesture interaction area of ​​interest, and if so, retain it; if not, discard it.

[0068] S32: Cluster the three-dimensional spatial position coordinates processed by S31, and determine whether the number of clusters obtained is greater than 0. If so, retain the cluster with the largest number of data points, discard other data points, and proceed to S33. If not, return to step S31.

[0069] S33: Calculate the mean position of the three-dimensional coordinates of all data points in the cluster, and calculate the distance from all data points to the mean position, and determine whether it is less than the set threshold. If so, retain it and proceed to S34. If not, discard it.

[0070] S34: Calculate the target parameters of the current frame using all the data points filtered by steps S31, S32 and S33, including but not limited to the number of point clouds, radial distance, radial velocity, azimuth, pitch angle, acceleration, scattering amplitude, three-dimensional rectangular coordinates, etc.

[0071] S35: Send the target parameters of the current frame into a sliding window of a certain frame length to form a gesture feature.

[0072] S36: Use some target parameters in the sliding window to determine whether there is a gesture. If so, proceed to step S4; if not, return to S1.

[0073] S4: Use a rule matching algorithm without machine learning and template library to directly determine which type of gesture feature the gesture feature conforms to and determine the gesture category (at this time, S4 does not need to collect a large number of gesture samples in advance to train the machine learning classification model, and does not need to set the gesture template library in advance. It only needs to use the set gesture judgment rules to determine which type of gesture feature the point cloud feature obtained in step S3 conforms to, thereby determining the gesture category). Using a rule matching algorithm to directly judge gesture features simplifies the model training process and reduces dependence on data volume.

[0074] Optionally, S4 includes the following steps:

[0075] S41: using the target parameter of each frame in the sliding window to determine whether the frame is a gesture frame;

[0076] S42: defining the first gesture frame in the sliding window as a starting gesture frame, and calculating the three-dimensional position of the starting gesture frame as the average of the three-dimensional positions of all data points in the frame as the starting gesture position;

[0077] S43: Calculate the mean of the three-dimensional positions of all data points in all gesture frames except the starting gesture frame as the central gesture position;

[0078] S44: Define the phase center of the two-dimensional antenna array of the radar as the origin of the three-dimensional rectangular coordinate system, define the azimuth, pitch and antenna array direction as the X direction, Z direction and Y direction respectively, calculate the difference between the center gesture position and the starting gesture position in the three coordinate directions, and the direction corresponding to the maximum difference is the gesture movement direction, define the +X direction gesture and the -X direction gesture as the left gesture and the right gesture respectively, define the +Z direction gesture and the -Z direction gesture as the upward gesture and the downward gesture respectively, and define the +Y direction gesture and the -Y direction gesture as the forward push gesture and the backward pull gesture respectively.

[0079] Optionally, the above S44 can identify Figure 3 The six types of gestures shown, including (a) upward, (b) downward, (c) left, (d) right, (e) pull back, and (f) push forward, enhance practicality.

[0080] S5: Display the output results of feature extraction and gesture recognition to achieve real-time visual interaction. By displaying the results of feature extraction and gesture recognition, real-time interaction between people and the system is achieved, improving the user experience.

[0081] like Figure 4 As shown, this embodiment also provides a radar gesture recognition system 100, which includes:

[0082] The radar subsystem 110 is used to transmit a frequency modulated continuous wave signal to a measurement area, receive an echo signal scattered from the measurement area, and convert the analog echo signal into a digital signal that can be processed by a data preprocessing module.

[0083] The data preprocessing subsystem 120 is used to process the digital signal to generate original radar point cloud data containing the gesture target.

[0084] The feature extraction subsystem 130 is used to filter the effective radar point cloud data, generate point cloud features, and determine whether a gesture exists.

[0085] The gesture identification subsystem 140 uses a rule matching algorithm without machine learning and template library to directly determine which type of gesture feature the point cloud feature matches and determine the gesture category.

[0086] The human-computer interaction subsystem 150 is used to display the output results of the feature extraction module and the gesture recognition module, and to realize the real-time visual interaction between human gestures and the radar gesture recognition system.

[0087] This modular system includes radar subsystem, data preprocessing subsystem, feature extraction subsystem, gesture recognition subsystem and human-computer interaction subsystem. Each module has distinct functions and is easy to maintain and expand. The data preprocessing, feature extraction and gesture recognition modules are integrated in the same hardware unit, which improves the compactness and efficiency of the system.

[0088] like Figure 5 As shown, based on the same inventive concept, this embodiment also provides a radar gesture recognition device 200, which includes a radar RF front-end module 210, a data preprocessing module 220, a feature extraction module 230, a gesture identification module 240, a human-computer interaction module 250, a communication module 260, a power module 270, and a housing module 280.

[0089] Among them, the radar RF front-end module 210 is placed in front of the observed gesture, transmits a frequency modulated continuous wave signal to the measurement area, receives the echo signal scattered by the measurement area, and converts the simulated echo signal into a digital signal that can be processed by the data preprocessing module; the data preprocessing module 220 processes the digital signal to generate original radar point cloud data containing gesture targets; the feature extraction module 230 screens the valid radar point cloud data, generates point cloud features, and determines whether there is a gesture; the gesture discrimination module 240 uses a rule matching algorithm without machine learning and template library to directly determine which point cloud feature meets the requirements. The human-computer interaction module 250 displays the output results of the feature extraction module and the gesture identification module, and realizes the real-time visual interaction between human gestures and the radar gesture recognition system. The communication module 260 is used to realize the communication between the modules inside the device and the communication with the outside of the device. The power module 270 is used to supply power to each module. The housing module 280 is used to install the radar RF front-end module 210, the data preprocessing module 220, the feature extraction module 230, the gesture identification module 240, the human-computer interaction module 250, the communication module 260, and the power module 270.

[0090] In this embodiment, the data preprocessing module 220, the feature extraction module 230 and the gesture identification module 240 are integrated into the same hardware unit, and the power module adopts a rechargeable lithium battery.

[0091] like Figure 6As shown, the radar RF front-end module 210 mainly includes a radar signal transmitting component 211, a radar signal receiving component 212, a mixer 213 and an analog-to-digital conversion component 214, etc. The radar signal transmitting component transmits a frequency modulated continuous wave signal to the measurement area, the radar signal receiving component collects the echo signal scattered by the measurement area, and the mixer and the analog-to-digital conversion component convert the analog echo signal into digital echo data.

[0092] like Figure 7 As shown, the electronic device in this embodiment includes a microcontroller unit (MCU), a radar radio frequency front end (RFFront-End), a random access memory (RAM), a bus (Bus) and an input / output (I / O) interface. The MCU can perform various defined processes according to the program instructions stored in the RAM, including but not limited to processing digital signals from the radar radio frequency front end module 210, executing the data preprocessing module 220, executing the feature extraction module 230, executing the gesture recognition module 240, internal communication and external communication, etc. The MCU, RF Front-End, RAM and I / O interface are all connected to each other through the bus. Radar radio frequency front end module: includes transmitting, receiving, mixing and analog to digital conversion components to ensure the efficiency of signal processing. And the microcontroller unit (MCU) is used to perform various processing tasks, which is flexible and scalable.

[0093] The storage medium in this embodiment stores program instructions for execution by the MCU.

[0094] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0095] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0096] In the context of the present invention, specific examples of storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. Supporting multiple storage media ensures the effective storage and execution of program code, and the program code can be executed in multiple environments, enhancing the adaptability of the system.

[0097] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A radar gesture recognition method, characterized in that: The following steps are involved: S1, transmitting a frequency modulated continuous wave signal to a measurement area, receiving an echo signal scattered from the measurement area, and converting the analog echo signal into a digital signal containing gesture information; S2, performing data processing on the digital signal containing the gesture information to generate original radar point cloud data containing the gesture target; S3: Filter valid radar point cloud data, generate point cloud features, and determine whether there is a gesture. If yes, proceed to S4, if not, return to S1; S4: Use a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the gesture features match and determine the gesture category; S5: Display the output results of feature extraction and gesture recognition to achieve real-time visual interaction.

2. The radar gesture recognition method according to claim 1, characterized in that: The S3 specifically includes the following steps: S31, determining whether each data point in the original radar point cloud of the current frame is within the gesture interaction area of ​​interest; if so, retaining it; if not, discarding it; S32, clustering the three-dimensional spatial position coordinates processed by S31, and determining whether the number of clusters obtained is greater than 0; if so, retaining the cluster with the largest number of data points, discarding other data points, and proceeding to S33; if not, returning to S31; S33, calculating the mean position of the three-dimensional coordinates of all data points in the cluster, and calculating the distance from all data points to the mean position, and judging whether it is less than the set threshold; if so, retain it and proceed to S34, if not, discard it; S34, calculating the target parameters of the current frame using all the data points filtered by S31, S32 and S33; S35, sending the target parameters of the current frame into a sliding window of a certain frame length to form a gesture feature; S36. Use some target parameters in the sliding window to determine whether there is a gesture; if so, proceed to S4; if not, return to S1.

3. The radar gesture recognition method according to claim 1, characterized in that: The S4 specifically includes the following steps: S41. Determine whether each frame is a gesture frame by using the target parameters of the frame in the sliding window. S42, defining the first gesture frame in the sliding window as a starting gesture frame, and calculating the three-dimensional position of the starting gesture frame as the average of the three-dimensional positions of all data points contained in the frame as the starting gesture position; S43, calculating the mean of the three-dimensional positions of all data points in all gesture frames except the starting gesture frame as the central gesture position; S44. Define the phase center of the two-dimensional antenna array of the radar as the origin of the three-dimensional rectangular coordinate system, define the azimuth, pitch and antenna array direction as the X direction, Z direction and Y direction respectively, calculate the difference between the center gesture position and the starting gesture position in the three coordinate directions, and the direction corresponding to the maximum difference is the gesture movement direction; define the +X direction gesture and the -X direction gesture as the left gesture and the right gesture respectively, define the +Z direction gesture and the -Z direction gesture as the upward gesture and the downward gesture respectively, and define the +Y direction gesture and the -Y direction gesture as the forward push gesture and the backward pull gesture respectively.

4. A radar gesture recognition system, characterized in that: include: Radar subsystem: used to transmit frequency modulated continuous wave signals to the measurement area, receive echo signals scattered from the measurement area, and convert analog echo signals into digital signals that can be processed by the data preprocessing subsystem; Data preprocessing subsystem: used to process digital signals and generate original radar point cloud data containing gesture targets; Feature extraction subsystem: used to filter effective radar point cloud data, generate point cloud features, and determine whether there is a gesture; Gesture identification subsystem: Uses a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the point cloud features match and determine the gesture category; Human-computer interaction subsystem: used to display the output results of the feature extraction subsystem and gesture recognition subsystem, and realize real-time visual interaction between human gestures and radar gesture recognition system.

5. A radar gesture recognition device, characterized in that: include: The radar RF front-end module is placed in front of the observation gesture, transmits a frequency modulated continuous wave signal to the measurement area, receives the echo signal scattered from the measurement area, and converts the analog echo signal into a digital signal that can be processed by the data preprocessing module; The data preprocessing module processes the digital signal and generates the original radar point cloud data containing the gesture target; Feature extraction module, which filters valid radar point cloud data, generates point cloud features, and determines whether there is a gesture; The gesture recognition module uses a rule matching algorithm without machine learning and template library to directly determine which type of gesture features the point cloud features match and determine the gesture category; The human-computer interaction module displays the output results of the feature extraction module and the gesture recognition module, and realizes the real-time visual interaction between human gestures and the radar gesture recognition device.

6. The radar gesture recognition device according to claim 5, characterized in that: The radar RF front-end module includes a radar signal transmitting component, a radar signal receiving component, a mixer and an analog-to-digital conversion component; The radar signal transmitting component transmits a frequency modulated continuous wave signal to the measurement area, the radar signal receiving component collects the echo signal scattered from the measurement area, and the mixer and the analog-to-digital conversion component convert the analog echo signal into digital echo data.

7. The radar gesture recognition device according to claim 5, characterized in that: The radar RF front-end module includes at least one transmitting antenna and three receiving antennas, and the three receiving antennas are arranged in an L-shaped array, and azimuth and elevation angle measurements are simultaneously achieved, thereby obtaining a target point cloud in three-dimensional space; Or at least two transmitting antennas and two receiving antennas are included to realize an equivalent two-dimensional antenna array.

8. The radar gesture recognition device according to claim 5, characterized in that: The signal transmitted by the radar RF front-end module is a frequency modulated continuous wave system.

9. An electronic device, comprising a processing unit, a display unit and a communication unit, characterized in that: The processing unit implements the method according to any one of claims 1 to 3 when executing the executable program; The display unit is used to display the intermediate process and the final result; The communication unit is used for communication within the system and communication with the outside of the system.

10. The electronic device according to claim 9, characterized in that: It also includes a storage unit, wherein the storage medium stores a feasible program, and the executable program is executed by the processing unit according to any one of claims 1-3.

Images