Human Presence Detection Method, Device, and Storage Medium Based on Doppler Radar

Abstract

This invention provides a method, device, and storage medium for human presence detection based on Doppler radar. The method includes: preprocessing intermediate frequency signals sampled within the sensing range; determining whether a moving object exists in the current space based on human motion detection signals; performing time-domain signal processing and analysis on micro-motion and respiratory frequency range signals to determine whether human micro-motion or respiratory frequency exists in the current space; converting the preprocessed time-domain signal into a frequency-domain signal, and extracting the human respiratory frequency signal through processing and spectral analysis of the frequency-domain signal to determine whether a human body exists in the current space; and determining the presence status of a human body within the current sensing range based on human motion detection, time-domain respiratory signal detection, and frequency-domain respiratory signal detection. This invention can effectively filter interference signals and improve the accuracy and sensitivity of human presence detection.

Description

Technical Field

[0001] This disclosure relates to the field of sensing technology, and in particular to a method, device, and storage medium for human presence sensing and detection based on Doppler radar. Background Technology

[0002] With the development of IoT technology, various IoT devices are becoming increasingly widespread. To improve the intelligence level of these devices, various sensors are needed, and Doppler radar sensors are one such type. These sensors enable smart devices to detect human movement, subtle movements, and even respiratory and heart rate signals within a certain space, accurately determining whether someone is present. Doppler radar sensors are widely used in energy-saving lighting, smart homes, and security monitoring due to their advantages such as good penetration and insensitivity to external ambient temperature.

[0003] Traditional respiratory and heart rate detection methods are primarily contact-based, requiring the device to be worn on the body to complete the detection. This method limits the scope of application; for example, it cannot detect the presence of someone in a specific space, and the user experience is poor because contact is required. There are also radar-based methods for respiratory detection, typically using the following two approaches:

[0004] 1. Hardware filtering is used to process radar signals. This requires adding hardware amplification circuits, low-pass filter circuits, and high-pass filter circuits to filter out signals other than respiratory signals step by step. Then, the intermediate frequency signal is acquired and the amplitude is judged to determine whether a human body is present.

[0005] 2. Convert the sampled radar intermediate frequency time domain signal into a frequency domain signal using FFT. Find the frequency corresponding to the maximum energy by calculating the signal energy value within 0 to 1 Hz, which is the breathing frequency.

[0006] Hardware filtering increases circuit design complexity, making it unsuitable for various complex application scenarios, lacking versatility and flexibility, and increasing hardware costs. Existing methods, such as FFT, typically process Doppler signals across the entire frequency band to analyze the presence of people in the environment. However, this method itself is susceptible to interference, increasing the difficulty of breath detection and is time-consuming, hindering the use of smart devices in various scenarios. Summary of the Invention

[0007] In view of this, the present disclosure provides a human presence detection method, device, and storage medium based on Doppler radar. The detection method is based on a Doppler radar sensor chip and uses software to implement a digital filter to filter out interference signals other than breathing and heart rate. It combines multiple methods such as time domain signal analysis and frequency domain signal analysis to detect whether there is a person in the current environment, which can effectively filter interference signals and improve the accuracy and sensitivity of human presence detection.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A human presence detection method based on Doppler radar, comprising:

[0010] The intermediate frequency signal sampled within the sensing range is preprocessed to obtain the processed motion detection signal, micro-motion and respiratory frequency range signal;

[0011] The motion detection signal is used to determine whether there is a moving object in the current space, and the human motion detection result is obtained.

[0012] Time-domain signal processing and analysis are performed on the micro-movement and respiratory frequency range signals to determine whether there are human micro-movements or respiratory frequencies in the current space, and the time-domain respiratory signal detection results are obtained.

[0013] The time-domain signal after signal preprocessing is converted into a frequency-domain signal. By processing and spectral analysis of the frequency-domain signal, the human breathing frequency signal is extracted to determine whether there is a human body in the current space, and the frequency-domain breathing signal detection result is obtained.

[0014] Based on the human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results, the existence status of the human body within the current sensing range is determined.

[0015] Furthermore, the sampled intermediate frequency signal includes intermediate frequency data sampled by the radar chip.

[0016] Furthermore, the signal preprocessing employs a digital low-pass filter, digital downsampling, and a digital stopband filter to downconvert the original intermediate frequency signal and filter out interference signals.

[0017] Further, it includes: obtaining the motion detection signal through a sliding window motion object detection algorithm to determine whether there is a moving object in the current space.

[0018] Furthermore, time-domain signal processing and analysis are performed on the micro-motion and respiratory frequency range signals to determine whether there are human micro-motions or respiratory frequencies present in the current space. Specifically, this includes...

[0019] The micro-movement and respiratory frequency range signals are divided into multiple windows using a windowing mechanism. The signal amplitude difference within each window is calculated and compared with a preset time-domain respiratory signal amplitude threshold to determine whether there is micro-movement or respiratory frequency.

[0020] Furthermore, the pre-processed time-domain signal is converted into a frequency-domain signal. By processing and performing spectral analysis on this frequency-domain signal, it is determined whether a human body is present in the current space. Specifically, this includes...

[0021] The preprocessed data is converted from the time domain signal to the frequency domain signal by performing FFT (Fast Fourier Transform). The frequency domain signal is then subjected to spectrum analysis to filter out interference signals and extract the energy value of the respiratory signal frequency range. This energy value is then compared with a preset threshold to determine whether someone is present.

[0022] Furthermore, the method for obtaining the energy value includes: obtaining the energy of the corresponding points of the frequency range within 0.1Hz-0.6Hz by calculating the amplitude of the frequency domain data after performing FFT (Fast Fourier Transform), and calculating the sum of the amplitudes of each frequency range within 0.1Hz-0.6Hz to obtain the energy value.

[0023] Furthermore, if any one or more of the human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results indicate the presence of a human body, then it is determined that a human body exists within the current sensing range.

[0024] The present invention also provides a human presence sensing and detection device based on Doppler radar, including a Doppler radar sensor and a processor, wherein the processor implements the steps of the above-described human presence sensing and detection method based on Doppler radar when executing a stored computer program.

[0025] The present invention also provides a storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described Doppler radar-based human presence sensing detection method.

[0026] The present invention relates to a Doppler radar-based human presence detection method, device, and storage medium. The detection method first performs signal preprocessing on the sampled intermediate frequency signal to filter out out-of-band interference signals and power frequency and its harmonic interference signals. The signal is processed in multiple stages. At the same time, the method combines time domain signal analysis and frequency domain signal analysis to effectively filter various interference signals and accurately detect whether there is a person in the current environment, thereby improving the accuracy and sensitivity of human presence detection. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a basic functional block diagram of the sensing detection method in one embodiment of the present invention;

[0029] Figure 2 This is a flowchart of the software implementation of the sensing detection method in one embodiment of the present invention;

[0030] Figure 3 This is a software flowchart of a sliding window moving object detection method in one embodiment of the present invention. Detailed Implementation

[0031] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0032] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0033] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using other structures and / or functionalities besides one or more of the aspects set forth herein.

[0034] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0035] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0036] This disclosure provides a Doppler radar-based human presence detection method, including:

[0037] The intermediate frequency signal sampled within the sensing range is preprocessed to obtain the processed motion detection signal, micro-motion and respiratory frequency range signal;

[0038] The motion detection signal is used to determine whether there is a moving object in the current space, and the human motion detection result is obtained.

[0039] Time-domain signal processing and analysis are performed on the micro-movement and respiratory frequency range signals to determine whether there are human micro-movements or respiratory frequencies in the current space, and the time-domain respiratory signal detection results are obtained.

[0040] The time-domain signal after signal preprocessing is converted into a frequency-domain signal. By processing and spectral analysis of the frequency-domain signal, the human breathing frequency signal is extracted to determine whether there is a human body in the current space, and the frequency-domain breathing signal detection result is obtained.

[0041] Based on the human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results, the existence status of the human body within the current sensing range is determined.

[0042] This invention is based on motion detection using Doppler radar. The principle is as follows: Doppler radar emits a fixed-frequency pulse signal to scan the air. When the electromagnetic wave signal encounters a moving object in the air, the reflected echo signal will produce a certain frequency difference with the transmitted signal, i.e., Doppler frequency shift. The radar module mixes the received echo signal and the transmitted signal to obtain the corresponding intermediate frequency signal. The intermediate frequency signal is filtered to remove interference signals, and then analyzed by a specific algorithm to obtain information such as the target object's speed, size, distance, and trajectory.

[0043] Doppler radar sensors can be used in energy-saving lamps, smart homes, and security monitoring. They can detect signals from large moving objects and, through specific algorithms, process subtle signals from relatively stationary individuals, even respiratory and heart rate signals. Human respiration causes chest and abdominal movements; a normal human respiratory rate is 0.1Hz-0.6Hz. Doppler radar can capture these respiratory movements, and after processing with specific algorithms, it can extract the respiratory signal, thus determining whether a person is present in a given space.

[0044] like Figure 1 As shown, Figure 1 This is a basic functional block diagram of a detection method according to an embodiment of the present invention. The detection method in this embodiment is based on the AT5820 5.8GHz Doppler radar sensor chip developed by GeKong Intelligent Technology Co., Ltd. It uses software to implement a digital filter to filter out interference signals other than breathing and heart rate. It combines multiple methods, including time-domain signal analysis and frequency-domain signal analysis, to detect the presence of a person in the current environment, effectively filtering interference signals and improving the accuracy and sensitivity of human presence detection.

[0045] The software processing mainly includes radar intermediate frequency sampling, signal preprocessing, human body micro-motion detection, time domain signal processing and analysis, time domain breathing signal decision, frequency domain signal processing and analysis, frequency domain breathing signal decision, and human body micro-motion decision.

[0046] (1) Radar intermediate frequency sampling 01

[0047] The AT5820 chip acquires radar intermediate frequency data through its built-in ADC and saves the data in a designated cache via DMA transfer.

[0048] (2) Signal preprocessing 02

[0049] Signal preprocessing includes digital low-pass filters, digital downsampling, and digital stopband filters to downconvert the original intermediate frequency signal, filter out out-of-band signals and power frequency and its harmonic interference signals, and acquire signals in the range of motion detection, micro-motion detection and breathing frequency.

[0050] (3) Human motion detection 03

[0051] After the above signal preprocessing, most of the interference signals can be filtered out, and the sliding window moving object detection algorithm can be used to determine whether there are moving objects in the current space.

[0052] (4) Time-domain signal processing and analysis 04

[0053] After the aforementioned signal preprocessing, a large signal bandwidth is still retained. If it is necessary to detect respiratory signals, the signal needs to be further downsampled and a low-pass filter is used to filter out data other than respiratory signals, retaining only the signal in the normal human respiratory frequency range of 0.1Hz-0.6Hz.

[0054] (5) Time-domain-based micro-motion and respiratory signal judgment 05

[0055] After processing in step (4), the frequency range signal of human breathing is retained. The time domain signal is divided into several windows by a windowing mechanism. The signal amplitude difference within each window is calculated and compared with the preset time domain breathing signal amplitude threshold to determine whether there is micro-movement or breathing frequency.

[0056] (6) Frequency Domain Signal Processing and Analysis 06

[0057] The data after signal preprocessing in step (2) is converted from time domain signal to frequency domain signal by FFT fast Fourier transform. The frequency domain signal is then subjected to spectrum analysis to filter out interference signals and extract the energy value of the respiratory signal frequency range.

[0058] (7) Decision based on frequency domain respiratory signal 07

[0059] The presence of a person is determined by comparing the energy value of the respiratory signal frequency range extracted in step (6) with a preset threshold.

[0060] (8) Judgment on the Existence of the Human Body 08

[0061] By combining the above-mentioned human motion detection, time-domain respiratory signal judgment, and frequency-domain respiratory signal judgment, it can be finally determined whether there is a person in the current space.

[0062] Human motion detection can detect large human movements with high real-time performance. Time-domain-based respiratory micro-motion detection can detect minute movements such as typing or turning pages. Frequency-domain-based respiratory detection can detect respiratory signals when a person is sitting still or lying down to sleep. If any one or more of the above human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results indicate the presence of a human body, then it is determined that a human body exists within the current sensing range.

[0063] like Figure 2 As shown, Figure 2 This is a flowchart illustrating the software implementation of a detection method according to an embodiment of the present invention. The software flowchart is explained below:

[0064] (1) Radar intermediate frequency sampling 1

[0065] The radar chip's internal ADC samples the radar's intermediate frequency data and stores it in a cache via DMA for subsequent signal processing. The sampling frequency is set to 635Hz.

[0066] (2) Low-pass filter (130Hz) 2

[0067] In step (1), sampling is performed at 635Hz. The effective signal acquisition range is within 317Hz. The signal range of human movement is usually within 150Hz. In order to filter high-frequency interference signals, including power frequency and its multiple harmonic signals, this embodiment uses a 130Hz low-pass filter to filter the original sampled data. The processed data retains the effective signal below 130Hz.

[0068] (3) Digital downconversion 3

[0069] To facilitate subsequent signal processing and reduce the amount of data processed, the software uses a second-order CIC cascaded comb filter to downconvert the 635Hz signal to a 317.5Hz sampling frequency, while filtering out some higher frequency signals.

[0070] (4) Power frequency filtering processing 4

[0071] Interference signals of 50Hz / 60Hz and their harmonics of 100Hz / 120Hz will also exist within 130Hz. Band-stop filters of 50Hz / 60Hz and 100Hz / 120Hz were designed to filter out these interference signals.

[0072] (5) Time-domain signal analysis

[0073] After the above signal processing, most of the interference signals can be filtered out, and some abrupt signals can be filtered out when analyzing the characteristics of time-domain signal data, including amplitude and range of variation.

[0074] (6) Motion detection judgment 6

[0075] The sliding window motion object detection algorithm determines whether there is an object moving in the current space. The motion detection decision can detect whether there is a person moving in the space, including walking, moving, etc. The motion detection decision has high real-time performance. The entire process from data sampling to signal processing and decision can be completed and the results output within 200ms.

[0076] (7) Low-pass filter (30Hz) 7

[0077] For both time-domain and frequency-domain respiratory signal processing, only low-frequency signals need to be retained. Here, a 30Hz low-pass filter is used to filter out signals outside 30Hz.

[0078] (8) Digital downconversion 8

[0079] To facilitate subsequent respiratory signal processing and reduce the amount of data processed, a 13th-order CIC filter was used to downconvert the signal with a sampling rate of 317.5Hz to a sampling frequency of 24.4Hz.

[0080] (9) Low-pass filter (0.6Hz)

[0081] After the aforementioned signal preprocessing, a large signal bandwidth is still retained. If it is necessary to detect the respiratory signal, the signal needs to be further downsampled and a low-pass filter is used to filter out data other than the respiratory signal, retaining only the signal in the normal human respiratory frequency range of 0.1Hz-0.6Hz. Here, a 0.6Hz low-pass filter is used to filter out other signals.

[0082] (10) Time-domain signal analysis

[0083] Human respiration causes chest and abdominal movements. A normal human's respiratory rate is 0.1Hz-0.6Hz. Doppler radar can detect these respiratory movements, and after processing with specific algorithms, respiratory signals can be extracted to determine the presence of a person in a given space. The time-domain analysis method for respiratory signals primarily uses a sliding window-based moving object detection method to determine the presence of a person in the current space.

[0084] (11) Micro-motion detection decision 11

[0085] The signal amplitude value detected by the sliding window is compared with a preset threshold. When the amplitude exceeds the threshold, it is considered that someone is present.

[0086] (12) Fast Fourier Transform (FFT)

[0087] The data processed above is converted from a time-domain signal to a frequency-domain signal using Fast Fourier Transform (FFT). Before performing the FFT, to prevent signal energy from spreading to adjacent frequency points (i.e., spectral leakage), the signal needs to be windowed. A Hamming window is typically used, and its function is as follows:

[0088] W(n,α)=(1-α)-αcos(2*PI*n / (N-1)), 0≦n≦N-1

[0089] The windowed data is processed into FFT frames with 256 or 512 points as a single frame.

[0090] (13) Spectrum Analysis

[0091] The energy of the corresponding points in the frequency domain data after FFT is performed can be obtained by calculating the amplitude. The amplitudes in the frequency range of 0.1Hz-0.6Hz can be accumulated, or the frequency of the point corresponding to the largest amplitude can be taken as the human respiratory rate.

[0092] (14) Breath test judgment 14

[0093] The sum of amplitudes in the frequency range of 0.1Hz-0.6Hz is obtained by using the process (13) and compared with a preset threshold. If the sum exceeds the preset threshold, it is determined that there is a human breathing signal.

[0094] (15) Judgment on the Existence of the Human Body

[0095] The human presence determination process uses motion detection in step (6), micro-motion detection in step (11), and respiration detection in step (14) to logically determine the final result. Motion detection in step (6) can detect whether someone is moving in space, including walking and exercising. It offers high real-time performance, with the entire cycle from data sampling to signal processing and determination completed within 200ms. Micro-motion detection in step (11) can detect micro-motions such as typing, writing, and even breathing. The micro-motion detection cycle is between 3 and 10 seconds. Respiration detection in step (14) is based on FFT followed by spectral analysis, allowing for more accurate extraction of respiration signals.

[0096] In the above process (6), the sliding window motion object detection algorithm is used to determine whether there is an object moving in the current space. The methods used for motion detection decision and micro-motion detection decision are both based on the sliding window motion object detection method.

[0097] The software flowchart for the sliding window moving object detection method is as follows: Figure 3 As shown in the figure, the specific details are as follows:

[0098] 1) Set N points as a window, and collect the intermediate frequency data of N points as a window of data.

[0099] 2) Obtain the maximum and minimum values ​​in a frame of data, and take the difference as the amplitude difference of a window of waveform data.

[0100] 3) Compare the difference obtained from the above steps with the set threshold. If it is greater than the set threshold, increment the number of moving object detection markers by 1. Otherwise, jump to step 7) to prepare for the judgment of the next window.

[0101] 4) Increase the number of times the object moves by 1.

[0102] 5) Determine if the number of times the object moves is greater than the preset value. If it is greater than the preset value, send a signal that the object is moving to the upper layer for processing. Otherwise, jump to step 7) to prepare for the judgment of the next window.

[0103] 6) Send a signal indicating that an object is moving to the upper layer for processing and jump to step 9) to prepare for the judgment of the next cycle.

[0104] 7) Increment the number of windows in the statistics by 1.

[0105] 8) Determine whether the number of statistical windows is greater than the preset maximum statistical window value. If it exceeds the maximum statistical window number, clear the current moving object detection cycle and re-enter the next moving object detection cycle. If it is less than the maximum statistical window number, directly enter the next moving object detection cycle.

[0106] 9) Clear the parameters of the current moving object detection cycle and re-enter the next moving object detection cycle.

[0107] This invention improves the real-time performance and accuracy of human presence detection by combining three methods: moving object detection, micro-motion detection, and respiration detection. Various filters, including digital low-pass filters, CIC cascaded integrator-comb filters, and band-stop filters, effectively filter out-of-band interference signals, power frequency and harmonic interference signals, and progressively reduce the frequency of the raw data, decreasing the amount of data for subsequent signal processing and improving the accuracy of motion / micro-motion / respiration detection. Furthermore, the sliding window moving object detection method can filter out some burst signals, avoiding false triggers. The filters used in this invention can be implemented using other filters such as passband filters and notch filters. The parameters of the filters can be dynamically adjusted; for example, low-pass filtering can be set to below 0.6Hz, or even below 0.5Hz or 1Hz, and the parameter settings for other filters are similar.

[0108] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. A method for detecting the presence of a human body based on Doppler radar, characterized in that, Used to detect micro-motion signals and human respiratory and heart rate signals when a person is relatively stationary, including: The intermediate frequency signal sampled within the sensing range is preprocessed. The preprocessing includes sequential digital low-pass filtering, digital downsampling, and digital stopband filtering to filter out out-of-band interference signals and power frequency and harmonic interference signals, and to obtain the processed motion detection signal, micro-motion and breathing frequency range signal. The motion detection signal is processed by a sliding window motion object detection algorithm to determine whether there is a moving object in the current space, and the human motion detection result is obtained. The micro-movement and respiratory frequency range signals are processed and analyzed in the time domain. The signals are divided into multiple windows through a windowing mechanism. The signal amplitude difference within each window is calculated and compared with a preset time domain respiratory signal amplitude threshold to determine whether there is human micro-movement or respiratory frequency in the current space, and the time domain respiratory signal detection result is obtained. The time-domain signal after signal preprocessing is converted into a frequency-domain signal through Fast Fourier Transform (FFT). This frequency-domain signal is then processed and subjected to spectral analysis to filter out interference signals and extract energy values ​​within the 0.1Hz–0.6Hz frequency band. These values ​​are then compared with a preset frequency-domain respiratory signal energy threshold to extract the human respiratory frequency signal and determine whether a human body is present in the current space, thus obtaining the frequency-domain respiratory signal detection result. The method for obtaining the energy value includes: calculating the amplitude of the frequency-domain data after the FFT to obtain the energy at corresponding points within the 0.1Hz–0.6Hz frequency range; and summing the amplitudes within the 0.1Hz–0.6Hz frequency range to obtain the energy value. Based on the human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results, the existence status of the human body within the current sensing range is determined.

2. The human presence detection method based on Doppler radar according to claim 1, characterized in that, The sampled intermediate frequency signal includes intermediate frequency data sampled by the radar chip.

3. The human presence detection method based on Doppler radar according to claim 1, characterized in that, If any one or more of the human motion detection results, time-domain respiratory signal detection results, and frequency-domain respiratory signal detection results indicate the presence of a human body, then it is determined that a human body exists within the current sensing range.

4. A human presence sensing and detection device based on Doppler radar, comprising a Doppler radar sensor and a processor, characterized in that, When the processor executes the stored computer program, it implements the steps of the Doppler radar-based human presence sensing detection method according to any one of claims 1-3.

5. A storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the Doppler radar-based human presence sensing detection method according to any one of claims 1-3.

Images