A radar data processing method and radar data acquisition system

Abstract

The present invention provides a radar data processing method and a radar data acquisition system. It is applied to the field of medical radar. The method includes: initializing and configuring the radar parameters of the radar data transceiver, and obtaining the radar echo signal; extracting the vital sign information and body motion information according to the radar echo signal, and judging whether there are vital signs in the monitoring scene; continuously obtaining the judgment result of whether there are vital signs, if it is continuously displayed that there are no vital signs and the time exceeds the preset time, then adjusting the radar parameters to the low power consumption mode, and obtaining the radar echo signal around the bed in the low power consumption mode; performing body motion detection and body target integration according to the radar echo signal around the bed, and judging whether there are vital signs around the bed; continuously obtaining the judgment result of whether there are vital signs around the bed, if it is continuously displayed that there are vital signs and the time exceeds the preset time, then adjusting the radar parameters to the initialization state. The present invention reduces power consumption and optimizes the battery life, storage and data processing efficiency of the system.

Description

Technical Field

[0001] The present invention relates to the field of medical radar, and in particular to a radar data processing method and a radar data acquisition system. Background Art

[0002] With the development of technology, more and more attention to sleep quality has become an important trend in health management. Modern technology provides a variety of advanced sleep monitoring tools and technologies that can more accurately analyze sleep patterns and quality. These technologies not only help doctors better diagnose and treat sleep disorders, but also enable individuals to track their sleep in real time, so as to take effective measures to improve sleep quality.

[0003] Existing sleep monitoring solutions mainly include contact and non-contact types. Contact monitoring such as polysomnography (PSG) requires recording bioelectric signals through multi-channel lead wires. Although the data is detailed, the equipment is complex and needs to be worn by the patient; non-contact solutions save the trouble of wearing, but there are problems in operation and data processing. When non-contact sleep monitoring is manually set using buttons, the inconvenience of patient operation may lead to monitoring failure, and the automatic recording method is easy to collect invalid data when the patient is out of bed, wasting equipment battery and storage space.

[0004] Therefore, both contact and contactless devices face the challenge of data processing and are unable to effectively distinguish the patient's bedside status and process invalid data, which affects monitoring efficiency and increases costs. Summary of the invention

[0005] In view of this, the present invention provides a radar data processing method, comprising:

[0006] Initialize and configure radar parameters of the radar data transceiver of the radar data acquisition system;

[0007] Acquire a radar echo signal in a radar monitoring scene, and extract vital sign information and body movement information according to the radar echo signal;

[0008] Determine whether there are vital signs in the monitoring scene based on the vital sign information and body movement information, and continuously obtain the judgment result of whether there are vital signs in the monitoring scene;

[0009] If the judgment result continuously shows that there are no vital signs and the time exceeds the preset time, the radar parameters are adjusted to the low power consumption mode so that the radar monitoring scene is reduced to the bedside, and the radar echo signal around the bed in the low power consumption mode is obtained;

[0010] Performing body motion detection and body target integration according to the bed perimeter radar echo signal;

[0011] Determine whether there are vital signs around the bed based on the body motion detection results and body target integration results, and continuously obtain the determination results of whether there are vital signs around the bed;

[0012] If the judgment result continues to show the presence of vital signs and the time exceeds the preset time, the radar parameters are adjusted to the initialization state and the radar echo signal processing continues.

[0013] Optionally, the radar parameters include bandwidth and modulation time, wherein the bandwidth in low power mode is the same as the initialized bandwidth, and the modulation time in low power mode is less than the initialized modulation time, so that the radar data acquisition system reduces the signal detection distance in the low power mode while increasing the frequency modulation slope and reducing the number of sampling points in the radar echo signal and increasing the time interval between the radar echo signals.

[0014] Optionally, extracting vital sign information according to the radar echo signal includes:

[0015] Performing a range-dimensional fast Fourier transform on the radar echo signal;

[0016] Accumulate multiple radar echo signals within the cycle time, and perform velocity-dimensional fast Fourier transform to obtain Doppler phase information within the cycle time;

[0017] Performing de-wrapping processing on the Doppler phase information to obtain a small speed change signal of the target in a fixed area;

[0018] The slight speed change signal is processed in time domain and frequency domain respectively to obtain the user's vital sign information.

[0019] Optionally, extracting body motion information according to the radar echo signal includes:

[0020] Extracting the radar echo signal;

[0021] Performing range-dimensional fast Fourier transform on the extracted radar echo signal;

[0022] Perform clutter cancellation processing on radar echo signals after range-dimensional fast Fourier transform;

[0023] Perform moving target recognition on the radar echo signal after clutter processing to obtain the target position point;

[0024] According to the change of the target position point, the user's body movement information is extracted.

[0025] Optionally, judging whether there are vital signs in the monitoring scene according to the vital sign information and the body movement information includes:

[0026] Constructing the point cloud of the target position according to the vital signs information and body motion information;

[0027] The constructed point cloud is condensed into an actual object model, and users and interference sources are identified based on the actual object model;

[0028] If the actual object model is a user, determining that there are vital signs in the monitoring scene;

[0029] If the actual object model is an interference source, it is determined that no vital signs exist in the monitoring scene.

[0030] Optionally, after the vital sign information is extracted or the judgment result continuously shows the existence of vital signs and the time exceeds a preset time, it also includes: filtering and deciding the current vital sign information according to the historical vital sign information, and packaging the filtered vital sign information for storage and uploading.

[0031] Optionally, performing body motion detection according to the bed periphery radar echo signal includes:

[0032] Performing a distance-dimensional fast Fourier transform on the radar echo signal around the bed, and performing a velocity-dimensional fast Fourier transform on multiple radar echo signals accumulated within a cycle time to obtain Doppler phase information around the bed within the cycle time;

[0033] estimating a large velocity component according to Doppler phase information around the bed;

[0034] Determining whether the large velocity component estimate exceeds a preset threshold value;

[0035] If the large speed component continues to be higher than the preset threshold value, the speed threshold determination device is triggered.

[0036] Optionally, performing volume target integration according to the bed periphery radar echo signal includes:

[0037] Performing a range-dimensional fast Fourier transform on the bed-surrounding radar echo signal, and performing clutter cancellation processing on the bed-surrounding radar echo signal after the range-dimensional fast Fourier transform;

[0038] Perform moving target recognition on the radar echo signal around the bed after clutter processing to obtain the moving target scattered points around the bed;

[0039] Aggregating the target scatter points around the bed, analyzing the motion trajectories of the aggregated target scatter points around the bed, and combining the time series data with dynamic feature extraction to obtain the user's body movement information around the bed;

[0040] Constructing a point cloud of a target position around the bed according to the body motion information around the bed;

[0041] The constructed point cloud is condensed into the actual object model around the bed to obtain the volume target.

[0042] Optionally, judging whether there are vital signs around the bed based on the body motion detection result and the body target integration result includes:

[0043] The body target is screened according to the breakthrough speed threshold decision device to determine whether there are dynamic characteristics consistent with vital signs around the bed;

[0044] If there are dynamic characteristics consistent with vital signs around the bed, the duration will be monitored. If it exceeds the specified duration, it is determined that there are vital signs around the bed;

[0045] If there are no dynamic features consistent with vital signs around the bed, it is determined that there are no vital signs around the bed.

[0046] A second aspect of the present invention provides a radar data acquisition system, which includes: a radar data transceiver, a central processing unit, an external device, a memory and a communication device, wherein the radar data transceiver is used to send radar signals and collect radar echo signals, the central processing unit is used to execute the radar data processing method, the memory is used to store filtered vital sign information packages, and the communication device is used to upload the stored data to the cloud.

[0047] Optionally, the central processor is further used to reduce the main frequency of the internal clock and shut down the external device when adjusting the radar parameters to the low power consumption mode;

[0048] The central processor is also used to restore the main frequency of the internal clock to the maximum and turn on the external device when adjusting the radar parameters to the initialization state.

[0049] The present invention achieves significant power consumption optimization and data processing efficiency improvement by intelligently adjusting the working mode of the radar data acquisition system. The system configures radar parameters and obtains echo signals in the initialization phase, and then determines whether there are vital signs by extracting vital signs and body movement information. When it is detected that there is no one for a long time, the system will switch to low-power mode, focusing on the area around the bed, reducing power consumption and storage requirements. When vital signs are detected and the time exceeds the preset value, the system will return to the initial state for comprehensive monitoring to obtain more and more accurate vital sign details. This method reduces power consumption and optimizes the system's battery life, storage and data processing efficiency by dynamically switching between two sets of signal processing algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 is a structural diagram of a radar data acquisition system in an embodiment of the present invention;

[0052] Figure 2 is a flowchart of a radar data processing method in an embodiment of the present invention;

[0053] Figure 3 A radar parameter configuration waveform design diagram for initialization in an embodiment of the present invention;

[0054] Figure 4 A radar parameter configuration waveform design diagram for a low power consumption mode in an embodiment of the present invention;

[0055] Figure 5 is a specific flow chart of the radar data processing method in an embodiment of the present invention;

[0056] Figure 6 is a flow chart of data processing in an embodiment of the present invention;

[0057] Figure 7 This is a flow chart of the automatic data compression and low power consumption mechanism in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments 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.

[0059] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0060] like Figure 1 As shown, an embodiment of the present invention provides a radar data acquisition system, including: a radar data transceiver 1, a central processing unit 2, an external device 3, a memory 4 and a communication device 5, the radar data transceiver 1 is used to send radar signals and collect radar echo signals, the central processing unit 2 is used to execute the radar data processing method, the memory 4 is used to package and store filtered vital sign information, and the communication device 5 is used to upload the stored data to the cloud.

[0061] Among them, the central processing unit is represented by MCU, and the communication equipment uploads the stored data to the cloud through communication protocols such as Wifi or BLE. The external devices are managed and controlled by the central processing unit, including indicator lights, network configuration buttons, USB interfaces, etc.

[0062] The radar data transceiver is an active millimeter-wave radar sensor. The sensor's transmitting antenna emits pulsed electromagnetic waves at a fixed period. The objects being measured (such as people moving indoors or patients undergoing sleep monitoring in bed) continuously reflect the electromagnetic waves and propagate the reflected electromagnetic waves to the receiving antenna of the millimeter-wave radar sensor. After the receiving antenna obtains the difference frequency signal through the low-noise amplifier and mixer output, the signal is filtered and sampled, and output to the central processor for signal processing.

[0063] The signals reflected back by targets with different characteristics vary greatly. A person walking in the room will reflect the continuous changes in the target's spatial position and the changes in movement speed. The rise and fall of the chest skin of a person lying on the bed due to breathing and heartbeat will be completely recorded. These signals will be processed and judged in real time by the central processing unit (MCU). Important data will be uploaded, and invalid data will be blocked.

[0064] like Figure 2 As shown, the central processing unit executes a radar data processing method, which specifically includes:

[0065] S1, initializing and configuring radar parameters of the radar data transceiver of the radar data acquisition system.

[0066] After the radar data acquisition system is powered on, the millimeter-wave radar chip is initialized and configured through the SPI interface of the central processor; the monitoring scene is the spatial area where the object to be measured is resting, and the distance range can be set.

[0067] S2, obtain the radar echo signal in the radar monitoring scene, and extract the vital sign information and body movement information according to the radar echo signal. The vital sign can be a patient or a caregiver, etc. The vital sign information includes vital sign data such as breathing and heartbeat, and the body movement information includes large-scale movements such as turning over, leg movement, and hand movement.

[0068] S3, judging whether there are vital signs in the monitoring scene according to the vital sign information and body movement information, and continuously obtaining the judgment result of whether there are vital signs in the monitoring scene. If the judgment result continuously shows that there are no vital signs and the time exceeds the preset time, execute step S4.

[0069] S4, adjusting the radar parameters to a low power consumption mode so that the radar monitoring scene is reduced to the periphery of the bed, and obtaining the radar echo signal of the periphery of the bed in the low power consumption mode.

[0070] When there are no vital signs for a long time in the monitoring scene, the radar parameters are adjusted to reduce power consumption. This configuration is sufficient to monitor the human body data around the bed. The monitoring scene is focused on the area around the patient's bed, and no longer pays attention to interference far away from the bed, but only pays attention to whether the patient has left the bed.

[0071] If the judgment result continues to show that there are vital signs and the time exceeds the preset time, the process returns to continue to execute the operation of step S2.

[0072] S5, body motion detection and body target integration are performed based on the radar echo signal around the bed.

[0073] S6, judging whether there are vital signs around the bed according to the body motion detection results and the body target integration results, and continuously obtaining the judgment result of whether there are vital signs around the bed. If the judgment result continuously shows that there are vital signs and the time exceeds the preset time, execute step S7.

[0074] S7, adjust radar parameters.

[0075] When vital signs appear around the bed, that is, the patient is in bed, continuous monitoring is performed. When the time in bed exceeds the preset time, in order to collect accurate patient data, the radar parameters are adjusted to the initialization state to monitor more human body data, and return to step S1 to continue monitoring the data.

[0076] This embodiment achieves significant power consumption optimization and data processing efficiency improvement by intelligently adjusting the working mode of the radar data acquisition system. The system configures radar parameters and obtains echo signals in the initialization phase, and then determines whether there are vital signs by extracting vital signs and body movement information. When it is detected that there is no one for a long time, the system will switch to low-power mode, focusing on the area around the bed to reduce power consumption and storage requirements. When vital signs are detected and the time exceeds the preset value, the system will return to the initial state for comprehensive monitoring to obtain more and more accurate vital sign details. This method reduces power consumption and optimizes the system's battery life, storage, and data processing efficiency by dynamically switching between two sets of signal processing algorithms.

[0077] Furthermore, radar parameters include bandwidth and modulation time, wherein the bandwidth in low power mode is the same as the initialized bandwidth, and the modulation time in low power mode is less than the initialized modulation time, thereby enabling the radar data acquisition system to reduce the signal detection distance in low power mode while increasing the frequency modulation slope and reducing the number of sampling points in the radar echo signal and increasing the time interval between radar echo signals.

[0078] The initial radar parameters are as follows: Figure 3 As shown in the figure, the bandwidth of the configured chirp signal is about 3GHz, and the modulation time is 214us. The high signal bandwidth is to bring finer distance resolution to the radar system, and the long modulation time is to obtain a larger range of target monitoring area. The chirp signal is repeated 32 times in 20ms. As long as the mode is not switched, the signal is sent and received repeatedly.

[0079] Radar parameters in low power mode are as follows: Figure 4As shown in the figure, compared with the initial radar parameter configuration, the bandwidth of this mode remains unchanged, but the frequency modulation slope is increased several times by reducing the frequency modulation time to 48us, and the number of sampling points in the chirp signal is reduced. At the same time, the time interval between chirp signals is increased by 2ms, and the number of chirps in each group of signals is reduced from the original 32 chirp signals to 4.

[0080] Although the radar configuration in low-power mode reduces the maximum distance range that the radar can monitor, it focuses more details on the area around the bed, no longer paying attention to interference far away from the bed, and only paying attention to whether the out-of-bed state is broken. Although this configuration slightly reduces the maximum speed that the radar can monitor, it is impossible for a person to have a very high speed after approaching the bed, so this configuration is sufficient to monitor larger human movements around the bed, such as sitting, lying, standing and walking after a person approaches the bed. The waveform design greatly reduces the sensitivity of speed detection, and micro-movements such as heartbeat and breathing will definitely not be detected under this radar configuration; but this mode is mainly a low-power design, and there is no need to pay too much attention to the patient's detailed vital signs.

[0081] like Figure 6 As shown (signal processing part on the left), step S2 extracts vital sign information according to the radar echo signal, specifically including:

[0082] S21a, performing range-dimensional fast Fourier transform on the radar echo signal.

[0083] All radar echo signals (chirp signals) are used to call the MCU's algorithm resources to perform fast Fourier transform (FFT) in the distance dimension.

[0084] S22a, accumulating multiple radar echo signals within the cycle time, and performing velocity-dimensional fast Fourier transform to obtain Doppler phase information within the cycle time.

[0085] After accumulating 32 radar echo signals within 20ms, a fast Fourier transform is performed in the velocity dimension to obtain the Doppler phase information of the target at each position within the 20ms time.

[0086] S23a, performing detangling processing on the Doppler phase information to obtain a small speed change signal of the target in a fixed area.

[0087] After being processed by the unwrapping algorithm, slight fluctuations in a certain fixed area will be reflected. For example, the present invention focuses on chest changes and heartbeat vibrations within a range of about 1.2 meters to 1.5 meters from the radar data transceiver.

[0088] S24a, performing time domain processing and frequency domain processing on the slight speed change signal to obtain the user's vital sign information.

[0089] Time domain processing uses the peak counting method to roughly estimate the time domain changes and phase trends of the signal; frequency domain processing uses high-resolution Doppler fast Fourier transform to accurately estimate the frequency of breathing and heartbeat, such as the number of breaths (breathing BPM) or heartbeats (heartbeat BPM) per minute, as well as the strength of breathing and heartbeat.

[0090] This embodiment effectively captures the motion characteristics of the target in the distance and speed directions by performing fast Fourier transform of the radar echo signal in the distance dimension and the speed dimension. Then, through the unwrapping algorithm, the focus is on the small fluctuations of the distance radar data transceiver within a certain range, such as the expansion and contraction of the chest and the vibration of the heartbeat, so as to accurately reflect the physiological changes of the user. Finally, the extracted small speed change signal is processed in the time domain and frequency domain. The change of the signal and the phase trend are roughly estimated through the time domain processing, and the frequency and strength of the breathing and heartbeat are estimated with high resolution through the frequency domain processing, so as to obtain the detailed vital signs information of the user. In this way, the vital signs of the user can be monitored efficiently and accurately.

[0091] like Figure 6 As shown in the figure (signal processing part on the right), step S2 extracts body motion information according to the radar echo signal, specifically including:

[0092] S21b, extracting the radar echo signal, for example, extracting the chirp data by decimating 1 out of 20, thereby saving algorithm processing time.

[0093] S22b, performing range-dimensional fast Fourier transform on the extracted radar echo signal.

[0094] S23b, performing clutter cancellation processing on the radar echo signal after the range-dimensional fast Fourier transform.

[0095] Clutter cancellation processing aims to reduce or eliminate background clutter or noise in radar echo signals. These clutters usually mask the real target signal, making target identification difficult. Through cancellation processing, the signal-to-noise ratio of the target signal can be improved, making target detection and identification more accurate.

[0096] S24b, performing moving target recognition on the radar echo signal after clutter processing to obtain the target position point.

[0097] The information of the moving target is abstracted into points, and the points corresponding to these radar targets correspond one-to-one to the distance between the patient's body on the bed and the radar.

[0098] S25b, extracting the user's body movement information based on the change of the target position point. The change of the target position point reflects the patient's large movements such as turning over, leg movement, and hand movement. Objects outside the bed, such as moving nursing staff and patients slowly leaving the bed, will also be recorded truthfully.

[0099] After the system is powered on, the first radar parameter configuration is performed according to the parameters when the user is in bed, and the following steps are performed: Figure 6 The signal is processed in the steps in the above steps, and the radar mode is switched according to the processing results; the two modes are switched according to the judgment results, and the user is in bed according to the Figure 6 The signal processing is performed in the steps, and when the user is not in bed, parameter configuration and signal processing are performed in low power consumption mode to form a closed-loop process.

[0100] This embodiment focuses on the analysis and judgment of target movement, because the patient may go to the toilet or get out of bed during the monitoring process. This part of the data is not of concern to the doctor. If it is recorded truthfully, it will increase the pressure of subsequent data processing and take up the time of subsequent image judgment staff to eliminate invalid data. In addition, the signal is cluttered to reduce or eliminate the background clutter or noise in the radar echo signal. Then, the body movement information is extracted through the position point, and accurate action information can be extracted, which is helpful to track the patient's dynamics in real time.

[0101] Furthermore, if Figure 6 As shown, judging whether there are vital signs in the monitoring scene based on the vital sign information and body movement information includes:

[0102] The target position point cloud is constructed based on the vital sign information and body motion information. The vital sign information and long-term body motion information are comprehensively determined, that is, the set of all points is accumulated over a long period of time to construct the point cloud information.

[0103] The constructed point cloud is condensed into the actual object model, and the user and interference source are identified based on the actual object model. The user refers only to the patient, while the interference source refers to the caregiver or other personnel.

[0104] If the actual object model is a user, it is determined that there are vital signs in the monitoring scene.

[0105] If the actual object model is an interference source, it is determined that there is no vital sign in the monitoring scene, which means that the user has left the bed, and the process proceeds to step S4 to switch the configuration.

[0106] This embodiment can accurately determine whether there are vital signs in the monitoring scene through comprehensive analysis of vital sign information and body movement information, combined with point cloud construction and actual object model recognition.

[0107] Furthermore, if Figure 5As shown, after the vital sign information is extracted in step S2 or the judgment result in step S3 continuously shows the existence of vital signs and the time exceeds the preset time, step S2 also includes: filtering and deciding the current vital sign information according to the historical vital sign information, and packaging the filtered vital sign information for storage and uploading.

[0108] The extracted vital signs data such as respiratory BPM, heart rate BPM, respiratory phase, heart rate phase, patient stability, data confidence, etc. are filtered according to the historical data stored in the memory, and the current vital signs information is filtered according to the historical vital signs information, which can effectively remove noise and outliers and improve the accuracy of the data, and then judge the presence or absence of vital signs again, and package the current data for storage, and package it into a protocol that can be recognized by communication equipment and external devices and upload it to the cloud service.

[0109] This embodiment ensures the accuracy and stability of the data by extracting vital sign information and filtering and making decisions on it, while also performing data storage management. This process not only reduces misjudgments and false alarms, but also achieves long-term data accumulation, providing a reliable basis for intelligent decision-making and further filtering.

[0110] like Figure 5 As shown, after adjusting the radar parameters, it is best to include timestamp storage, specifically saving the time information of the current system to the peripheral memory of the central processing unit MCU to mark the beginning of this unmanned scene mode.

[0111] In one embodiment, step S5 performs body motion detection based on the bed periphery radar echo signal, specifically including:

[0112] S51a, performing a distance-dimensional fast Fourier transform on the radar echo signal around the bed, performing a velocity-dimensional fast Fourier transform on the multiple radar echo signals accumulated within the cycle time, and obtaining Doppler phase information around the bed within the cycle time.

[0113] S52a, estimating large velocity components based on Doppler phase information around the bed.

[0114] S53a, determining whether the large velocity component estimate exceeds a preset threshold value.

[0115] S54a, if the large speed component continues to be higher than the preset threshold value, the speed threshold breaking determiner is triggered.

[0116] Specifically, all the chirps of radar echo signals around the bed are first transformed in the distance dimension, and then the larger body movements within the monitoring range are estimated. If a target point with a larger velocity component is measured, it proves that a moving target has entered the vicinity of the bed. The implementation method is to first perform a velocity dimension fast Fourier transform on the accumulated chirps of radar echo signals around the bed, so as to analyze the Doppler characteristics of the signal and estimate the parameters of the larger velocity component. If there is a continuous accumulation of velocity components exceeding the set threshold, the speed threshold decision device is triggered. If it does not exceed the preset threshold value, the signal continues to be collected to estimate the large velocity component.

[0117] This embodiment detects body motion through the radar echo signal around the bed, specifically including Fourier transforming the echo signal, estimating the velocity component, and comparing it with the preset threshold value. Its function is to accurately detect body motion in the area around the bed: by analyzing the Doppler phase information and the large velocity component, it can timely detect whether there is a moving target entering the vicinity of the bed, thereby triggering the velocity threshold to extract body motion information with a larger velocity.

[0118] In one embodiment, step S5 performs volume target integration according to the radar echo signal around the bed, specifically including:

[0119] S51b, performing a range-dimensional fast Fourier transform on the bed-surrounding radar echo signal, and performing clutter cancellation processing on the bed-surrounding radar echo signal after the range-dimensional fast Fourier transform.

[0120] S52b, performing moving target recognition on the radar echo signal around the bed after clutter processing to obtain moving target scatter points around the bed.

[0121] S53b, aggregate the scattered points of target movement around the bed, analyze the motion trajectories of the aggregated scattered points of target movement around the bed, and obtain the user's body movement information around the bed by combining the time series data and dynamic feature extraction.

[0122] S54b, constructing a point cloud of target positions around the bed according to the body motion information around the bed.

[0123] S55b, condensing the constructed point cloud into the actual object model around the bed to obtain a volume target.

[0124] Specifically, this embodiment integrates the detected moving targets into a volume target to provide preprocessing for the subsequent comprehensive decision-making of the target movement trend. The implementation method is to first perform a distance-dimensional fast Fourier transform on all the bed-surrounding radar echo signal chirps, then perform clutter cancellation on the distance-dimensional fast Fourier transform data, output the identified moving target scatter plot, and then perform clustering processing based on the characteristics of each point to construct a point cloud. The constructed point cloud information is further subjected to point trace condensation processing to form a simplified volume target.

[0125] This embodiment performs a series of processing on the radar echo signal around the bed, including distance-dimensional fast Fourier transform, clutter cancellation, moving target recognition, scattered point aggregation and point cloud construction, and finally integrates the out-of-body target. The purpose of this process is to systematically process and integrate the detected moving target information, and provide clear pre-processed data for subsequent target motion trend analysis. In this way, the dynamic changes around the bed can be accurately identified and tracked.

[0126] Furthermore, judging whether there are vital signs around the bed based on the body motion detection results and body target integration results includes:

[0127] The body targets are screened according to the breakthrough speed threshold judgement device to determine whether there are dynamic characteristics consistent with vital signs around the bed.

[0128] If there are dynamic characteristics around the bed that are consistent with vital signs, the duration will be monitored. If it exceeds the specified duration, it is determined that there are vital signs around the bed.

[0129] If there are no dynamic features consistent with vital signs around the bed, it is determined that there are no vital signs around the bed.

[0130] Specifically, by integrating the results of body movement detection and body targets and entering into subsequent comprehensive decisions, the patient's trend of returning to bed can be judged more accurately. The data that has been judged as returning to bed will undergo bed residence analysis, and the characteristics of the body targets will be used to determine whether the patient is actually in bed within the cumulative time period.

[0131] This embodiment uses the body motion detection results and body target integration results, combined with speed threshold judgment and duration monitoring, to more accurately judge whether the patient returns to bed and stays in bed. This comprehensive decision-making method improves the accuracy of the system's judgment on the patient's return to bed behavior.

[0132] In one embodiment, when adjusting the radar parameters to the low power consumption mode in step S3, the central processing unit needs to reduce the main frequency of the internal clock and shut down the external devices.

[0133] The central processing unit will shut down most of the external devices and reduce the internal clock frequency of the central processing unit. In the subsequent steps of S3, the radar data acquisition system will no longer store and send data, and only perform low-speed signal processing on a very small amount of radar data.

[0134] In step S5, when the radar parameters are adjusted to the initialization state, the central processor also needs to restore the main frequency of the internal clock to the maximum and turn on the external device.

[0135] The central processing unit will turn on the external device again and restore the internal clock frequency of the central processing unit to the maximum. The radar data acquisition system will enter the standby state of storing and sending data, and will subsequently return to S1 to continue monitoring the data.

[0136] Specifically, Figure 5 After the low-power setting is performed, signal processing is performed, that is, body motion detection and body target integration are performed according to the radar echo signal around the bed, and then the presence of vital signs around the bed is continuously judged based on the body motion detection results and the body target integration results. When there are vital signs, an interrupt wake-up operation is executed, that is, the external device is reopened, and the internal clock frequency of the central processing unit is restored to the maximum, and the radar data acquisition system will enter the standby state for storing and sending data.

[0137] Figure 7 For the automatic data compression and low power consumption mechanism process, when the radar data is recorded, there is a state where the patient leaves the bed, and it can automatically search and identify the person, such as searching and identifying in the results of judging whether there are vital signs in steps S3 and S6. When it is recognized that there is someone in the current bed, high-speed data recording is turned on, and the system storage interrupt priority is set to high priority for vital sign data analysis and recording in the occupied state; when it is recognized that there is no one in the current bed, the system marks the bed leaving timestamp and only stores the bed leaving data, sets the system storage interrupt priority to low priority, closes most of the peripheral interfaces, sets the radar sensor to a low-speed transceiver mechanism, minimizes system power consumption, and changes the radar parameters so that the system runs in low power consumption mode.

[0138] In low power mode, it switches to unmanned recognition. If there is no one, the system will remain in low power mode. When a person is recognized, the system algorithm will be awakened and the high-speed data recording and storage process will be restarted.

[0139] The traditional radar mechanism is a single-mode transceiver mechanism, which results in the maximum power consumption. When there are no vital signs, the system needs to store massive data all the time, which not only increases the power consumption of the system, but also wastes precious storage resources. The automatic data compression and low power consumption mechanism designed in the present invention, through a reasonable process, not only ensures that the data is not lost when there are vital signs, but also adaptively identifies that there is no one, and pushes the system into a low power consumption mode, saving resources, increasing battery life, and reducing the complexity of subsequent signal processing.

[0140] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0141] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0144] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. A radar data processing method, characterized in that: include: Initialize and configure radar parameters of the radar data transceiver of the radar data acquisition system; Get the radar echo signal within the radar monitoring scene, Performing a distance-dimensional fast Fourier transform on the radar echo signal; accumulating multiple radar echo signals within a periodic time and performing a speed-dimensional fast Fourier transform to obtain Doppler phase information within the periodic time; performing an unwrapping process on the Doppler phase information to obtain a small speed change signal of a target within a fixed area; performing time domain processing and frequency domain processing on the small speed change signal to obtain the user's vital sign information; Extracting the radar echo signal, performing a distance-dimensional fast Fourier transform on the extracted radar echo signal, performing clutter cancellation processing on the radar echo signal after the distance-dimensional fast Fourier transform, performing moving target recognition on the radar echo signal after the clutter processing to obtain a target position point, and extracting the user's body motion information according to the change of the target position point; Determine whether there are vital signs in the monitoring scene based on the vital sign information and body movement information, and continuously obtain the judgment result of whether there are vital signs in the monitoring scene; If the judgment result continuously shows that there are no vital signs and the time exceeds the preset time, the radar parameters are adjusted to the low power consumption mode so that the radar monitoring scene is reduced to the bedside, and the bedside radar echo signal in the low power consumption mode is obtained; According to the radar echo signal around the bed, body motion detection and body target integration are performed, wherein the body motion detection includes performing a distance-dimensional fast Fourier transform on the radar echo signal around the bed, performing a speed-dimensional fast Fourier transform on multiple radar echo signals accumulated within a cycle time, obtaining Doppler phase information of the bed around the cycle time, estimating a large velocity component based on the Doppler phase information of the bed, and judging whether the large velocity component estimate exceeds a preset threshold value. If the large velocity component continues to be higher than the preset threshold value, a speed threshold determination device is triggered; body target integration ... speed threshold determination on the bed around the bed. Performing a range-dimensional fast Fourier transform on the radar echo signal, and performing clutter cancellation processing on the bed-surrounding radar echo signal after the range-dimensional fast Fourier transform, performing moving target recognition on the bed-surrounding radar echo signal after the clutter processing, obtaining scatter points of moving targets around the bed, aggregating the scatter points of moving targets around the bed, analyzing the motion trajectories of the aggregated scatter points of moving targets around the bed, combining time series data and dynamic feature extraction to obtain the user's bed-surrounding body motion information, and constructing a point cloud of the bed-surrounding target position according to the bed-surrounding body motion information; agglomerating the constructed point cloud with an actual object model around the bed to obtain a body target; Determine whether there are vital signs around the bed based on the body motion detection results and body target integration results, and continuously obtain the determination results of whether there are vital signs around the bed; If the judgment result continues to show the presence of vital signs and the time exceeds the preset time, the radar parameters are adjusted to the initialization state and the radar echo signal processing continues.

2. The method according to claim 1, characterized in that The radar parameters include bandwidth and modulation time, wherein the bandwidth in low power mode is the same as the initialized bandwidth, and the modulation time in low power mode is less than the initialized modulation time, so that the radar data acquisition system reduces the signal detection distance in the low power mode while increasing the frequency modulation slope and reducing the number of sampling points in the radar echo signal and increasing the time interval between the radar echo signals.

3. The method according to claim 1, characterized in that Determine whether there are vital signs in the monitoring scene based on the vital signs information and body movement information, including: Constructing the point cloud of the target position according to the vital signs information and body motion information; The constructed point cloud is condensed into an actual object model, and users and interference sources are identified based on the actual object model; If the actual object model is a user, determining that there are vital signs in the monitoring scene; If the actual object model is an interference source, it is determined that no vital signs exist in the monitoring scene.

4. The method according to claim 1, characterized in that: After the vital sign information is extracted or the judgment result continuously shows the existence of vital signs and the time exceeds the preset time, it also includes: filtering and deciding the current vital sign information according to the historical vital sign information, and packaging the filtered vital sign information for storage and uploading.

5. The method according to claim 1, characterized in that Based on the results of body motion detection and body target integration, determine whether there are vital signs around the bed, including: The body target is screened according to the breakthrough speed threshold decision device to determine whether there are dynamic characteristics consistent with vital signs around the bed; If there are dynamic characteristics consistent with vital signs around the bed, the duration will be monitored. If it exceeds the specified duration, it is determined that there are vital signs around the bed; If there are no dynamic features consistent with vital signs around the bed, it is determined that there are no vital signs around the bed.

Images