WO2021017307A1 - 非接触性心率检测方法、系统、设备及存储介质 - Google Patents

非接触性心率检测方法、系统、设备及存储介质 Download PDF

Info

Publication number
WO2021017307A1
WO2021017307A1 PCT/CN2019/118082 CN2019118082W WO2021017307A1 WO 2021017307 A1 WO2021017307 A1 WO 2021017307A1 CN 2019118082 W CN2019118082 W CN 2019118082W WO 2021017307 A1 WO2021017307 A1 WO 2021017307A1
Authority
WO
WIPO (PCT)
Prior art keywords
channel
frequency
target
pulse wave
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2019/118082
Other languages
English (en)
French (fr)
Inventor
王义文
郑权
王健宗
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Publication of WO2021017307A1 publication Critical patent/WO2021017307A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow
    • A61B5/024Measuring pulse rate or heart rate
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow
    • A61B5/024Measuring pulse rate or heart rate
    • A61B5/02444Details of sensor

Definitions

  • the embodiments of the present application relate to the field of heart rate measurement, and in particular, to a non-contact heart rate detection method, system, device, and storage medium.
  • Heart rate refers to the number of heartbeats per minute, which is one of the important parameters of human metabolism and functional activities. For example, the increase in resting heart rate is widely regarded as an independent risk factor for the detection of cardiovascular disease. Routine testing of resting heart rate is beneficial to the prevention and rehabilitation of cardiovascular diseases.
  • HR heart rate
  • RR respiratory rate
  • HR variability HR variability
  • HRV Heart rate variability
  • the purpose of the embodiments of the present application is to provide a non-contact heart rate detection method, system, device, and storage medium, which can make the calculation of the heart rate value and the absolute value of the true heart rate error within 3, thereby improving the accuracy of detection .
  • an embodiment of the present application provides a non-contact heart rate detection method, including:
  • the heart rate value is calculated according to the target frequency.
  • an embodiment of the present application also provides a non-contact heart rate detection system, including:
  • An acquisition module for acquiring multiple frames of face images in the video to be processed
  • An extraction module for extracting preset facial regions in the multi-frame face image
  • the selection module is used to obtain the pixel value matrix of the R channel, the G channel and the B channel of the preset face area, and select one of the R channel, the G channel and the B channel as the target channel;
  • An amplifying module configured to amplify the pixel value matrix corresponding to the target channel through Euler image magnification to obtain a cardiovascular pulse wave sequence
  • the analysis module is used to analyze the frequency waveform of the cardiovascular pulse wave sequence, and select the frequency value corresponding to the largest peak among the peaks of the frequency waveform as the target frequency;
  • the calculation module is used to calculate the heart rate value according to the target frequency.
  • an embodiment of the present application further provides a computer device, the computer device includes a memory and a processor, the memory stores computer-readable instructions that can run on the processor, and the computer When the readable instructions are executed by the processor, the following steps are implemented:
  • the heart rate value is calculated according to the target frequency.
  • the embodiments of the present application also provide a non-volatile computer-readable storage medium.
  • the non-volatile computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions may Is executed by at least one processor, so that the at least one processor executes the following steps:
  • the heart rate value is calculated according to the target frequency.
  • the non-contact heart rate detection method, system, device and storage medium provided by the embodiments of the application select the face part from the video, and use Euler image magnification algorithm to magnify the pixel value matrix to obtain the blood vessel pulse wave sequence.
  • the waveform of the vascular pulse wave sequence is analyzed, and the heart rate value is finally calculated; the embodiment of the present application can make the calculation of the heart rate and the absolute value of the true heart rate have an error within 3, which improves the accuracy of detection.
  • FIG. 1 is a flowchart of Embodiment 1 of a non-contact heart rate detection method according to an embodiment of this application.
  • Fig. 2 is a flowchart of step S100 in Fig. 1 of an embodiment of the application.
  • Fig. 3 is a flowchart of step S104 in Fig. 1 of the embodiment of the application.
  • Fig. 4 is a flowchart of step S104A3 in Fig. 1 of the embodiment of the application.
  • Fig. 5 is a flowchart of step S106 in Fig. 1 of the embodiment of the application.
  • Fig. 6 is a flowchart of step S106B in Fig. 1 of the embodiment of the application.
  • FIG. 7 is a schematic diagram of program modules of Embodiment 2 of the non-contact heart rate detection system according to the embodiment of the application.
  • FIG. 8 is a schematic diagram of the hardware structure of the third embodiment of the computer equipment of this application.
  • FIG. 1 there is shown a flow chart of the non-contact heart rate detection method in the first embodiment of the present application. It can be understood that the flowchart in this method embodiment is not used to limit the order of execution of the steps.
  • the following exemplarily describes the computer device 2 as the execution subject. details as follows.
  • Step S100 Obtain multiple frames of face images in the video to be processed.
  • step S100 includes:
  • Step S100A acquiring the face image information of each frame of image in the video information frame by frame according to time sequence
  • Step S100B Count the number of image frames of valid images, where the valid images are one or more images containing face image information in the video information;
  • step S100C when the number of image frames of the effective image is greater than a preset threshold, proceed to the next step to stop acquiring the face image information of each frame of the image in the video information.
  • the number of image frames of the effective image is greater than the preset threshold, it is preliminarily determined that there is a living body, otherwise it is determined that there is no living body. This step is used to determine whether the user to be tested has entered the shooting range.
  • Step S102 Extract a preset face region in the multi-frame face image.
  • the nose area and forehead area of the face image information can be selected as the preset face area.
  • the nose area and forehead area are richer in capillaries, thus having better heart rate detection. Effective and low noise interference.
  • the nose triangle area includes three feature points, and the forehead area includes four feature points.
  • the feature points of each area form a topological structure. These topological structures represent the graph model of these areas. Each feature point is the node of these topological structures. These nodes It is mainly extracted and filtered through scale-invariant feature transformation matching algorithm and node position relationship; then, refined feature extraction is performed on each node neighborhood, and the distance between nodes is calibrated, which forms the nose triangle area and forehead The feature point map of the four key areas.
  • Step S104 Obtain the pixel value matrix of the R channel, the G channel, and the B channel of the preset face area, and select one of the R channel, the G channel and the B channel as the target channel.
  • the area pixel value matrix of each pixel in the preset face area can be obtained through an image algorithm (such as the OpenCV algorithm).
  • the preset face area can be 50mm*50mm
  • the extracted pixel value matrix can be a 50*50*3 area pixel value matrix.
  • Each data in the area pixel value matrix is used to represent the corresponding pixel in a certain channel of each frame The pixel value of the point.
  • step S104 includes:
  • step S104A the pixel value matrix of each channel of the R channel, the G channel, and the B channel is transformed in the frequency domain through a fast Fourier transform to obtain the channel energy of each channel.
  • step S104A further includes:
  • Step S104A1 according to obtaining the area pixel value matrix of each of the R channel, G channel, and B channel in each frame;
  • Step S104A2 splicing the regional pixel value matrices of each of the R channel, G channel and B channel frame by frame to obtain a pixel value matrix
  • Step S104A3 calculate the channel energy of each channel.
  • the channel energy represents the change value of the pixel value matrix of each R channel, G channel, and B channel.
  • step S104A3 includes:
  • the pixel value matrix of each of the R channel, the G channel, and the B channel is transformed into the frequency domain through a fast Fourier transform:
  • the pixel value matrix of each channel is represented by x(n), and x(n) is decomposed into the sum of two sequences of even and odd numbers, namely:
  • x(n) x 1 (n)+x 2 (n);
  • the time length of x 1 (n) and x 2 (n) are both N/2, N represents the time length of each channel selection, x 1 (n) is an even sequence, x 2 (n) is an odd sequence, and then pixel
  • the value matrix performs fast Fourier transform operation, and the fast Fourier transform calculation formula is as follows:
  • the value of X(k) of each channel is calculated to obtain the channel energy.
  • step S104B the channel with the largest energy is selected as the target channel.
  • the channel corresponding to the maximum X(k) value is taken as the target channel, which indicates that the change value of the target channel is the largest.
  • Step S106 Enlarge the pixel value matrix corresponding to the target channel through Euler image enlargement to obtain a cardiovascular pulse wave sequence.
  • the embodiment of the present application amplifies the energy channel of the pixel value matrix of the target channel, and the signal-to-noise ratios of different basebands should be relatively close. Therefore, a Gaussian pyramid can be selected to down-sampling and low-pass filtering the target channel.
  • step S106 includes:
  • Step S106A spatially filtering the pixel value matrix corresponding to the target channel to obtain basebands of different spatial frequencies, wherein a low-pass filter is used for spatial filtering;
  • step S106B the baseband is smoothed and down-sampled according to the Gaussian pyramid to obtain the cardiovascular pulse wave sequence.
  • step S106B includes:
  • Step S106B1 the first-level Gaussian pyramid obtains the second-level Gaussian image through smoothing and down-sampling, and the cut-off frequency of the Gaussian pyramid gradually increases by a factor of 2 from the upper level to the next level;
  • Step S106B2 until the K-1 Gaussian pyramid is smoothed and down-sampling to obtain the K-th Gaussian image, and the cardiovascular pulse wave time sequence is obtained.
  • the pixel value matrix of the target channel after the Euler image is enlarged if it is necessary to reconstruct the pixel value matrix of the target channel after the Euler image is enlarged.
  • the pixel value matrix of the target channel is regarded as the smallest level of the baseband for down-sampling when the Euler image is enlarged
  • the R of the video to be processed The pixel value matrix of each channel in the channel, G channel and B channel can be superimposed to obtain the pixel value matrix of the target channel after Euler image enlargement, which can more obviously display the color change in the video under test.
  • Euler image amplification technology helps to reduce noise, and the image shows different signal-to-noise ratios at different spatial frequencies. Generally speaking, the lower the spatial frequency, the higher the signal-to-noise ratio. Therefore, to prevent distortion, these basebands should use different magnifications. For the topmost image, that is, the image with the lowest spatial frequency and the highest signal-to-noise ratio, the maximum magnification can be used, and the magnification of the next layer decreases in turn. Facilitate the approximation of the image signal.
  • Step S108 Analyze the frequency waveform of the cardiovascular pulse wave sequence, and select the frequency value corresponding to the largest peak among the peaks of the frequency waveform as the target frequency.
  • the frequency waveform of the cardiovascular pulse wave sequence is band-pass filtered, and the frequency range of the human heart rate is selected for band-pass filtering.
  • the band-pass filter can be a Butterworth band-pass filter or an ideal band-pass filter. Remove noise and other frequency domain interference to heart rate prediction. Different bandpass filters can be selected according to different needs. If you need to perform subsequent time-frequency analysis of the cardiovascular pulse wave sequence, you can choose the ideal bandpass filter; if you do not need to perform the time-frequency analysis of the cardiovascular pulse wave sequence, you can choose a filter with a wide passband, such as Butterworth Bandpass filter, secondary IIR filter, etc. The ideal low-pass filter is selected for this application.
  • step S108 includes:
  • the frequency bandwidth 0.4-4Hz (24-240bpm) is selected as the analysis frequency band, and the frequency waveform of the cardiovascular pulse wave sequence is band-pass filtered to obtain the peaks of the frequency waveforms of the cardiovascular pulse wave sequence, wherein the maximum peak value corresponds to The frequency is the target frequency.
  • Step S110 calculating a heart rate value according to the target frequency.
  • the change of the face color comes from the blood change caused by the heartbeat, which can be measured from the heartbeat. It can be seen from the waveform of the frequency waveform of the vascular pulse wave sequence that the heart rate, the number of heartbeats per minute, is equal to 60 times the target frequency.
  • FIG. 7 shows a schematic diagram of the program modules of the second embodiment of the non-contact heart rate detection system of the present application.
  • the non-contact heart rate detection system 20 may include or be divided into one or more program modules.
  • the one or more program modules are stored in a storage medium and executed by one or more processors.
  • the above non-contact heart rate detection method can be realized.
  • the program module referred to in the embodiments of the present application refers to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment:
  • the acquiring module 200 is used to acquire multiple frames of face images in the video to be processed.
  • the obtaining module 200 is further used for:
  • the next step is to stop acquiring the face image information of each frame of the video information.
  • the number of image frames of the effective image is greater than the preset threshold, it is preliminarily determined that there is a living body, otherwise it is determined that there is no living body. This step is used to determine whether the user to be tested has entered the shooting range.
  • the extraction module 202 is configured to extract preset facial regions in the multi-frame face image.
  • the nose area and forehead area of the face image information can be selected as the preset face area.
  • the nose area and forehead area are richer in capillaries, thus having better heart rate detection. Effective and low noise interference.
  • the selecting module 204 is configured to obtain the pixel value matrix of the R channel, the G channel, and the B channel of the preset face region, and select one of the R channel, the G channel, and the B channel as the target channel.
  • the area pixel value matrix of each pixel in the preset face area can be obtained through an image algorithm (such as the OpenCV algorithm).
  • the preset face area can be 50mm*50mm
  • the extracted pixel value matrix can be a 50*50*3 area pixel value matrix.
  • Each data in the area pixel value matrix is used to represent the corresponding pixel in a certain channel of each frame The pixel value of the point.
  • the selection module 204 is also used for:
  • the pixel value matrix of each channel of the R channel, the G channel, and the B channel is transformed into the frequency domain through a fast Fourier transform to obtain the channel energy of each channel.
  • the channel energy represents the change value of the pixel value matrix of each R channel, G channel, and B channel.
  • the pixel value matrix of each of the R channel, the G channel, and the B channel is transformed into the frequency domain through a fast Fourier transform:
  • the pixel value matrix of each channel is represented by x(n), and x(n) is decomposed into the sum of two sequences of even and odd numbers, namely:
  • x(n) x 1 (n)+x 2 (n);
  • the time length of x 1 (n) and x 2 (n) are both N/2, N represents the time length of each channel selection, x 1 (n) is an even sequence, x 2 (n) is an odd sequence, and then the pixel
  • the value matrix performs fast Fourier transform operation, and the fast Fourier transform calculation formula is as follows:
  • the value of X(k) of each channel is calculated to obtain the channel energy.
  • the channel with the largest energy is selected as the target channel, that is, the channel corresponding to the maximum X(k) value is selected as the target channel.
  • the magnification module 206 is used to magnify the pixel value matrix corresponding to the target channel through Euler image magnification to obtain a cardiovascular pulse wave sequence.
  • the embodiment of the present application amplifies the energy channel of the pixel value matrix of the target channel, and the signal-to-noise ratios of different basebands should be relatively close. Therefore, a Gaussian pyramid can be selected to down-sampling and low-pass filtering the target channel.
  • the amplification module 206 is also used for:
  • the baseband is smoothed and down-sampled according to the Gaussian pyramid to obtain the cardiovascular pulse wave sequence.
  • the first-level Gaussian pyramid obtains the second-level Gaussian image through smoothing and down-sampling, and the cut-off frequency of the Gaussian pyramid gradually increases by a factor of 2 from the upper level to the next level;
  • the K level Gaussian image is obtained, and the cardiovascular pulse wave time sequence is obtained.
  • the pixel value matrix of the target channel after the Euler image is enlarged if it is necessary to reconstruct the pixel value matrix of the target channel after the Euler image is enlarged.
  • the pixel value matrix of the target channel is regarded as the smallest level of the baseband for down-sampling when the Euler image is enlarged
  • the R of the video to be processed The pixel value matrix of each channel in the channel, G channel and B channel can be superimposed to obtain the pixel value matrix of the target channel after Euler image enlargement, which can more obviously display the color change in the video under test.
  • Euler image amplification technology helps to reduce noise, and the image shows different signal-to-noise ratios at different spatial frequencies. Generally speaking, the lower the spatial frequency, the higher the signal-to-noise ratio. Therefore, to prevent distortion, these basebands should use different magnifications. For the topmost image, that is, the image with the lowest spatial frequency and the highest signal-to-noise ratio, the maximum magnification can be used, and the magnification of the next layer decreases in turn. Facilitate the approximation of the image signal.
  • the analysis module 208 is configured to analyze the frequency waveform of the cardiovascular pulse wave sequence, and select the frequency value corresponding to the largest peak among the peaks of the frequency waveform as the target frequency.
  • analysis module 208 is further used for:
  • the frequency bandwidth 0.4-4Hz (24-240bpm) is selected as the analysis frequency band, and the frequency waveform of the cardiovascular pulse wave sequence is band-pass filtered to obtain the peaks of the frequency waveforms of the cardiovascular pulse wave sequence, wherein the maximum peak value corresponds to The frequency is the target frequency.
  • the frequency waveform of the cardiovascular pulse wave sequence is band-pass filtered, and the frequency range of the human heart rate is selected for band-pass filtering.
  • the band-pass filter can be a Butterworth band-pass filter or an ideal band-pass filter. Remove noise and other frequency domain interference to heart rate prediction. Different bandpass filters can be selected according to different needs. If you need to perform subsequent time-frequency analysis of the cardiovascular pulse wave sequence, you can choose the ideal bandpass filter; if you do not need to perform the time-frequency analysis of the cardiovascular pulse wave sequence, you can choose a filter with a wide passband, such as Butterworth Bandpass filter, secondary IIR filter, etc. The ideal low-pass filter is selected for this application.
  • the calculation module 210 is configured to calculate a heart rate value according to the target frequency.
  • the change of the face color comes from the blood change caused by the heartbeat, which can be measured from the heartbeat. It can be seen from the waveform of the frequency waveform of the vascular pulse wave sequence that the heart rate, the number of heartbeats per minute, is equal to 60 times the target frequency.
  • the computer device 2 is a device that can automatically perform numerical calculation and/or information processing in accordance with pre-set or stored instructions.
  • the computer device 2 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server, or a server cluster composed of multiple servers).
  • the computer device 2 at least includes, but is not limited to, a memory 21, a processor 22, a network interface 23, and a non-contact heart rate detection system 20 that can communicate with each other through a system bus. among them:
  • the memory 21 includes at least one type of non-volatile computer-readable storage medium.
  • the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), Random access memory (RAM), static random access memory (SRAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk Wait.
  • the memory 21 may be an internal storage unit of the computer device 2, such as a hard disk or memory of the computer device 2.
  • the memory 21 may also be an external storage device of the computer device 2, for example, a plug-in hard disk, a smart media card (SMC), and a secure digital (Secure Digital, SD card, Flash Card, etc.
  • the memory 21 may also include both the internal storage unit of the computer device 2 and its external storage device.
  • the memory 21 is generally used to store the operating system and various application software installed in the computer device 2, for example, the program code of the non-contact heart rate detection system 20 in the second embodiment.
  • the memory 21 can also be used to temporarily store various types of data that have been output or will be output.
  • the processor 22 may be a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments.
  • the processor 22 is generally used to control the overall operation of the computer device 2.
  • the processor 22 is used to run the program code or process data stored in the memory 21, for example, to run the non-contact heart rate detection system 20 to implement the non-contact heart rate detection method of the first embodiment.
  • the network interface 23 may include a wireless network interface or a wired network interface.
  • the network interface 23 is generally used to establish a communication connection between the server 2 and other electronic devices.
  • the network interface 23 is used to connect the server 2 to an external terminal through a network, and to establish a data transmission channel and a communication connection between the server 2 and the external terminal.
  • the network may be Intranet, Internet, Global System of Mobile Communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G Network, Bluetooth (Bluetooth), Wi-Fi and other wireless or wired networks.
  • GSM Global System of Mobile Communication
  • WCDMA Wideband Code Division Multiple Access
  • 4G network Fifth Generation
  • 5G Network Fifth Generation
  • Bluetooth Bluetooth
  • Wi-Fi Wireless Fidelity
  • the non-contact heart rate detection system 20 stored in the memory 21 may also be divided into one or more program modules, and the one or more program modules are stored in the memory 21 and configured by One or more processors (the processor 22 in this embodiment) are executed to complete the application.
  • FIG. 7 shows a schematic diagram of the program modules of the second embodiment of the non-contact heart rate detection system 20.
  • the non-contact heart rate detection system 20 can be divided into an acquisition module 200 and an extraction module 202.
  • the selection module 204, the amplification module 206, the analysis module 208, and the calculation module 210 can be divided into an acquisition module 200 and an extraction module 202.
  • the selection module 204, the amplification module 206, the analysis module 208, and the calculation module 210 the program module referred to in this application refers to a series of computer program instruction segments that can complete specific functions.
  • the specific functions of the program modules 200-210 have been described in detail in the second embodiment, and will not be repeated here.
  • This embodiment also provides a non-volatile computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory ( SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, server, App application mall, etc., on which storage There are computer-readable instructions, and the corresponding functions are realized when the program is executed by the processor.
  • the non-volatile computer-readable storage medium of this embodiment is used to store the non-contact heart rate detection system 20, and when executed by the processor, the following steps are implemented:
  • the heart rate value is calculated according to the target frequency.
  • the channel with the largest energy is selected as the target channel.
  • the non-contact heart rate detection method, system, device and storage medium provided by the embodiments of the application select the face part from the video, and use Euler image magnification algorithm to magnify the pixel value matrix to obtain the blood vessel pulse wave sequence.
  • the waveform of the vascular pulse wave sequence is analyzed, and the heart rate value is finally calculated; the embodiment of the present application can make the calculation of the heart rate and the absolute value of the true heart rate have an error within 3, which improves the accuracy of detection.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Cardiology (AREA)
  • Biomedical Technology (AREA)
  • Medical Informatics (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Engineering & Computer Science (AREA)
  • Physiology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)

Abstract

一种非接触性心率检测系统、设备、存储介质及方法,该方法包括:获取待处理视频中的多帧人脸图像(S100);提取所述多帧人脸图像中的预设面部区域(S102);获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并选取其中一个通道作为目标通道(S104);通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列(S106);分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率(S108);根据目标频率计算得到心率值(S110)。该方法可以将心率值的计算与真实心率绝对值误差控制在3以内,提高了检测的精确性。

Description

非接触性心率检测方法、系统、设备及存储介质
本申请要求于2019年7月31日提交中国专利局,申请号、专利名称分别为201910699269.6、“非接触性心率检测方法、系统、设备及存储介质”发明专利的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及心率测量领域,尤其涉及一种非接触性心率检测方法、系统、设备及存储介质。
技术背景
心率是指每分钟心跳的次数,是人体新陈代谢以及功能活动重要的一个参数之一。举个例子,静息心率的增快就被广泛的认为是一项检测心血管疾病的独立危险因素,对静息心率进行日常检测有利于心血管疾病的预防以及康复治疗。
目前,在当今的生物医学研究领域,有很多种研究者提出了不同的非接触式测量生命体征,如心率(HR,heart rate)和呼吸频率(RR,respiratorn rate)的解决方案,其中包括激光多普勒,微波多普勒雷达和热成像方法。而HR变异的非接触式评估(HRV,Heart rate variability)是心脏自主神经活动的指标,更是一个更大的挑战。然而,发明人发现,有些研究尝试过去测评HRV,取得了令人印象深刻的进步,但是存在一个共同缺点是系统昂贵并需要专业硬件,并且电子硬件有寿命期限,会影响测试的准确度。
发明内容
有鉴于此,本申请实施例的目的是提供一种非接触性心率检测方法、系统、设备及存储介质,可以将心率值的计算与真实心率绝对值误差在3以内,提高 了检测的精确性。
为实现上述目的,本申请实施例提供了一种非接触性心率检测方法,包括:
获取待处理视频中的多帧人脸图像;
提取所述多帧人脸图像中的预设面部区域;
获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
根据所述目标频率计算得到心率值。
为实现上述目的,本申请实施例还提供了一种非接触性心率检测系统,包括:
获取模块,用于获取待处理视频中的多帧人脸图像;
提取模块,用于提取所述多帧人脸图像中的预设面部区域;
选取模块,用于获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
放大模块,用于通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
分析模块,用于分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
计算模块,用于根据所述目标频率计算得到心率值。
为实现上述目的,本申请实施例还提供了一种计算机设备,所述计算机设备包括存储器、处理器,所述存储器上存储有可在所述处理器上运行的计算机可读指令,所述计算机可读指令被所述处理器执行时实现以下步骤:
获取待处理视频中的多帧人脸图像;
提取所述多帧人脸图像中的预设面部区域;
获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
根据所述目标频率计算得到心率值。
为实现上述目的,本申请实施例还提供了一种非易失性计算机可读存储介质,所述非易失性计算机可读存储介质内存储有计算机可读指令,所述计算机可读指令可被至少一个处理器所执行,以使所述至少一个处理器执行以下步骤:
获取待处理视频中的多帧人脸图像;
提取所述多帧人脸图像中的预设面部区域;
获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
根据所述目标频率计算得到心率值。
本申请实施例提供的非接触性心率检测方法、系统、设备及存储介质,通过从视频中选取人脸部分,并利用欧拉影像放大算法对像素值矩阵进行放大处理得到血管脉搏波序列,再分析血管脉搏波序列的波形,最后计算出心率值;本申请实施例可以将心率的计算与真实心率绝对值误差在3以内,提高了检测的精确性。
附图说明
图1为本申请实施例非接触性心率检测方法实施例一的流程图。
图2为本申请实施例图1中步骤S100的流程图。
图3为本申请实施例图1中步骤S104的流程图。
图4为本申请实施例图1中步骤S104A3的流程图。
图5为本申请实施例图1中步骤S106的流程图。
图6为本申请实施例图1中步骤S106B的流程图。
图7为本申请实施例非接触性心率检测系统实施例二的程序模块示意图。
图8为本申请计算机设备实施例三的硬件结构示意图。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白,以下结合附图及实施例,对本申请进行进一步详细说明。应当理解,此处所描述的具体实施例仅用以解释本申请,并不用于限定本申请。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
实施例一
参阅图1,示出了本申请实施例一之非接触性心率检测方法的步骤流程图。可以理解,本方法实施例中的流程图不用于对执行步骤的顺序进行限定。下面以计算机设备2为执行主体进行示例性描述。具体如下。
步骤S100,获取待处理视频中的多帧人脸图像。
示例性的,参阅图2,步骤S100包括:
步骤S100A,依据时间顺序逐帧获取所述视频信息中的每帧图像的人脸图像信息;
步骤S100B,统计有效图像的图像帧数,所述有效图像为所述视频信息中含人脸图像信息的一个或多个图像;
步骤S100C,当所述有效图像的图像帧数大于预设阈值时,进入下一步则 停止获取所述视频信息中的每帧图像的人脸图像信息。
具体的,有效图像的图像帧数大于预设阈值时,则初步判断有活体存在,否则判断为无活体存在,该步骤用于确定待测用户是否进入拍摄范围内。
步骤S102,提取所述多帧人脸图像中的预设面部区域。
具体的,为避免使用整张人脸导致地噪声干扰,可以选取人脸图像信息的鼻子区域和额头区域作为预设面部区域,鼻子区域和额头区域毛细血管比较丰富,从而具有更好的心率检测效果,且噪音干扰小。鼻子三角区域包括三个特征点,额头区域包括四个特征点,各个区域的特征点形成拓扑结构,这些拓扑结构来表征这些区域的图模型,各个特征点便是这些拓扑结构的节点,这些节点主要通过尺度不变特征变换匹配算法和节点位置关系进行提取和筛选;然后,对每一个每个节点邻域进行精细化特征提取,节点之间的距离标定,这就形成了鼻子三角区域和额头的四个关键区域的特征点图。
步骤S104,获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道。
具体的,可以通过图像算法(如OpenCV算法)获取预设面部区域中各个像素点的区域像素值矩阵。预设面部区域可以为50mm*50mm,所提取的像素值矩阵可以为50*50*3的区域像素值矩阵,区域像素值矩阵中的每个数据用于表示每帧某个通道中的对应像素点的像素值。
示例性的,参阅图3,步骤S104包括:
步骤S104A,将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量。
进一步的,参阅图4,步骤S104A进一步包括:
步骤S104A1,根据获取每帧所述R通道、G通道和B通道中各个通道的区域像素值矩阵;
步骤S104A2,将各所述R通道、G通道和B通道的区域像素值矩阵逐帧 进行拼接得到像素值矩阵;
步骤S104A3,计算各个通道的通道能量。
具体的,通道能量表示各R通道、G通道和B通道的像素值矩阵的变化值。
示例性的,步骤S104A3包括:
将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换:
将各通道的像素值矩阵以x(n)表示,并将x(n)进行分解为偶数与奇数的两个序列之和,即:
x(n)=x 1(n)+x 2(n);
x 1(n)和x 2(n)的时间长度都是N/2,N表示各通道选取的时间长度,x 1(n)是偶数序列,x 2(n)是奇数序列,再对像素值矩阵进行快速傅里叶变换运算,所述快速傅里叶变换计算公式如下:
Figure PCTCN2019118082-appb-000001
计算各通道的X(k)的值得到所述通道能量。
步骤S104B,选择能量最大的通道作为所述目标通道。
具体的,将最大X(k)的值对应的通道作为目标通道,表示该目标通道的变化值最大。
步骤S106,通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列。
具体的,本申请实施例是对目标通道的像素值矩阵的能量通道进行放大,不同基带的信噪比应该比较接近,因此可以选择高斯金字塔,对目标通道进行下采样和低通滤波。
示例性的,参阅图5,步骤S106包括:
步骤S106A,将所述目标通道对应的像素值矩阵进行空间滤波,以得到不同的空间频率的基带,其中,采用低通滤波器进行空间滤波;
步骤S106B,根据高斯金字塔对所述基带进行平滑与下采样得到所述心血 管脉搏波序列。
示例性的,参阅图6,步骤S106B包括:
步骤S106B1,第一层高斯金字塔通过平滑与下采样获得二层高斯图像,高斯金字塔的截至频率从上一层到下一层以因子2逐渐增加;
步骤S106B2,直至第K-1层高斯金字塔通过平滑与下采样获得第K层高斯图像,得到所述心血管脉搏波时间序列。
具体的,若需重建经过欧拉影像放大后的目标通道的像素值矩阵。重建时,只需将高斯金字塔最小的一级进行上采样(因为进行欧拉影像放大时,将目标通道的像素值矩阵视为基带的最小级进行了下采样),最后与待处理视频的R通道、G通道和B通道中各个通道的像素值矩阵进行叠加就可以得到欧拉影像放大后的目标通道的像素值矩阵,能更明显的显示在待测视频中色彩的变化。
采用欧拉影像放大技术有助于减少噪声,图像在不同空间频率下呈现出不同的信噪比。一般来说,空间频率越低,信噪比反而越高。因此,为了防止失真,这些基带应该使用不同的放大倍数。最顶层的图像,即空间频率最低、信噪比最高的图像,可使用最大的放大倍数,下一层的放大倍数依次减小。便于对图像信号的逼近。
步骤S108,分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率。
具体的,将心血管脉搏波序列的频率波形进行带通滤波,选择人的心率的频段范围进行带通滤波,带通滤波器可以选用巴特沃斯带通滤波器或者理想带通滤波器,过滤掉噪声以及其他频域对心率预测的干扰。可以根据不同的需求选择不同的带通滤波器。如果需要对心血管脉搏波序列进行后续的时频分析,则可以选择理想带通滤波器;如果不需要对心血管脉搏波序列进行时频分析,可以选择宽通带的滤波器,如巴特沃斯带通滤波器,二级IIR滤波器等。本申请选用理想低通滤波器。
示例性的,步骤S108包括:
选择频率带宽0.4~4Hz(24~240bpm)作为分析频段,将所述心血管脉搏波序列的频率波形进行带通滤波得到所述心血管脉搏波序列的频率波形的波峰,其中波峰峰值最大对应的频率为所述目标频率。
步骤S110,根据所述目标频率计算得到心率值。
具体的,当相邻的多帧人脸图像的面部颜色差异大于预设阈值(即,待测用户的面部基本处于静止状态)时,其面部颜色变动来源于心跳引起的血液变化,可以从心血管脉搏波序列的频率波形的波形上看出,心率即每分钟心跳数等于60乘以目标频率。
实施例二
请继续参阅图7,示出了本申请非接触性心率检测系统实施例二的程序模块示意图。在本实施例中,非接触性心率检测系统20可以包括或被分割成一个或多个程序模块,一个或者多个程序模块被存储于存储介质中,并由一个或多个处理器所执行,以完成本申请,并可实现上述非接触性心率检测方法。本申请实施例所称的程序模块是指能够完成特定功能的一系列计算机程序指令段。以下描述将具体介绍本实施例各程序模块的功能:
获取模块200,用于获取待处理视频中的多帧人脸图像。
示例性的,所述获取模块200还用于:
依据时间顺序逐帧获取所述视频信息中的每帧图像的人脸图像信息;
统计有效图像的图像帧数,所述有效图像为所述视频信息中含人脸图像信息的一个或多个图像;
当所述有效图像的图像帧数大于预设阈值时,进入下一步则停止获取所述视频信息中的每帧图像的人脸图像信息。
具体的,有效图像的图像帧数大于预设阈值时,则初步判断有活体存在,否则判断为无活体存在,该步骤用于确定待测用户是否进入拍摄范围内。
提取模块202,用于提取所述多帧人脸图像中的预设面部区域。
具体的,为避免使用整张人脸导致地噪声干扰,可以选取人脸图像信息的鼻子区域和额头区域作为预设面部区域,鼻子区域和额头区域毛细血管比较丰富,从而具有更好的心率检测效果,且噪音干扰小。
选取模块204,用于获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道。
具体的,可以通过图像算法(如OpenCV算法)获取预设面部区域中各个像素点的区域像素值矩阵。预设面部区域可以为50mm*50mm,所提取的像素值矩阵可以为50*50*3的区域像素值矩阵,区域像素值矩阵中的每个数据用于表示每帧某个通道中的对应像素点的像素值。
示例性的,选取模块204还用于:
将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量。
进一步的,根据获取每帧所述R通道、G通道和B通道中各个通道的区域像素值矩阵;
将各所述R通道、G通道和B通道的区域像素值矩阵逐帧进行拼接得到像素值矩阵;
计算各个通道的通道能量。
具体的,通道能量表示各R通道、G通道和B通道的像素值矩阵的变化值。
示例性的,将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换:
将各通道的像素值矩阵以x(n)表示,并将x(n)进行分解为偶数与奇数的两个序列之和,即:
x(n)=x 1(n)+x 2(n);
x 1(n)和x 2(n)的时间长度都是N/2,N表示各通道选取的时间长度,x 1(n) 是偶数序列,x 2(n)是奇数序列,再对像素值矩阵进行快速傅里叶变换运算,所述快速傅里叶变换计算公式如下:
Figure PCTCN2019118082-appb-000002
计算各通道的X(k)的值得到所述通道能量。
选择能量最大的通道作为所述目标通道,即将最大X(k)的值对应的通道作为目标通道。
放大模块206,用于通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列。
具体的,本申请实施例是对目标通道的像素值矩阵的能量通道进行放大,不同基带的信噪比应该比较接近,因此可以选择高斯金字塔,对目标通道进行下采样和低通滤波。
示例性的,放大模块206还用于:
将所述目标通道对应的像素值矩阵进行空间滤波,以得到不同的空间频率的基带,其中,采用低通滤波器进行空间滤波;
根据高斯金字塔对所述基带进行平滑与下采样得到所述心血管脉搏波序列。
示例性的,第一层高斯金字塔通过平滑与下采样获得二层高斯图像,高斯金字塔的截至频率从上一层到下一层以因子2逐渐增加;
直至第K-1层高斯金字塔通过平滑与下采样获得第K层高斯图像,得到所述心血管脉搏波时间序列。
具体的,若需重建经过欧拉影像放大后的目标通道的像素值矩阵。重建时,只需将高斯金字塔最小的一级进行上采样(因为进行欧拉影像放大时,将目标通道的像素值矩阵视为基带的最小级进行了下采样),最后与待处理视频的R通道、G通道和B通道中各个通道的像素值矩阵进行叠加就可以得到欧拉影像放大后的目标通道的像素值矩阵,能更明显的显示在待测视频中色彩的变化。
采用欧拉影像放大技术有助于减少噪声,图像在不同空间频率下呈现出不同的信噪比。一般来说,空间频率越低,信噪比反而越高。因此,为了防止失 真,这些基带应该使用不同的放大倍数。最顶层的图像,即空间频率最低、信噪比最高的图像,可使用最大的放大倍数,下一层的放大倍数依次减小。便于对图像信号的逼近。
分析模块208,用于分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率。
示例性的,所述分析模块208还用于:
选择频率带宽0.4~4Hz(24~240bpm)作为分析频段,将所述心血管脉搏波序列的频率波形进行带通滤波得到所述心血管脉搏波序列的频率波形的波峰,其中波峰峰值最大对应的频率为所述目标频率。
具体的,将心血管脉搏波序列的频率波形进行带通滤波,选择人的心率的频段范围进行带通滤波,带通滤波器可以选用巴特沃斯带通滤波器或者理想带通滤波器,过滤掉噪声以及其他频域对心率预测的干扰。可以根据不同的需求选择不同的带通滤波器。如果需要对心血管脉搏波序列进行后续的时频分析,则可以选择理想带通滤波器;如果不需要对心血管脉搏波序列进行时频分析,可以选择宽通带的滤波器,如巴特沃斯带通滤波器,二级IIR滤波器等。本申请选用理想低通滤波器。
计算模块210,用于根据所述目标频率计算得到心率值。
具体的,当相邻的多帧人脸图像的面部颜色差异大于预设阈值(即,待测用户的面部基本处于静止状态)时,其面部颜色变动来源于心跳引起的血液变化,可以从心血管脉搏波序列的频率波形的波形上看出,心率即每分钟心跳数等于60乘以目标频率。
实施例三
参阅图8,是本申请实施例三之计算机设备的硬件架构示意图。本实施例中,所述计算机设备2是一种能够按照事先设定或者存储的指令,自动进行数 值计算和/或信息处理的设备。该计算机设备2可以是机架式服务器、刀片式服务器、塔式服务器或机柜式服务器(包括独立的服务器,或者多个服务器所组成的服务器集群)等。如图8所示,所述计算机设备2至少包括,但不限于,可通过系统总线相互通信连接存储器21、处理器22、网络接口23、以及非接触性心率检测系统20。其中:
本实施例中,存储器21至少包括一种类型的非易失性计算机可读存储介质,所述可读存储介质包括闪存、硬盘、多媒体卡、卡型存储器(例如,SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘等。在一些实施例中,存储器21可以是计算机设备2的内部存储单元,例如该计算机设备2的硬盘或内存。在另一些实施例中,存储器21也可以是计算机设备2的外部存储设备,例如该计算机设备2上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。当然,存储器21还可以既包括计算机设备2的内部存储单元也包括其外部存储设备。本实施例中,存储器21通常用于存储安装于计算机设备2的操作系统和各类应用软件,例如实施例二的非接触性心率检测系统20的程序代码等。此外,存储器21还可以用于暂时地存储已经输出或者将要输出的各类数据。
处理器22在一些实施例中可以是中央处理器(Central Processing Unit,CPU)、控制器、微控制器、微处理器、或其他数据处理芯片。该处理器22通常用于控制计算机设备2的总体操作。本实施例中,处理器22用于运行存储器21中存储的程序代码或者处理数据,例如运行非接触性心率检测系统20,以实现实施例一的非接触性心率检测方法。
所述网络接口23可包括无线网络接口或有线网络接口,该网络接口23通常用于在所述服务器2与其他电子装置之间建立通信连接。例如,所述网络接口23用于通过网络将所述服务器2与外部终端相连,在所述服务器2与外部终端之间的建立数据传输通道和通信连接等。所述网络可以是企业内部网 (Intranet)、互联网(Internet)、全球移动通讯系统(Global System of Mobile communication,GSM)、宽带码分多址(Wideband Code Division Multiple Access,WCDMA)、4G网络、5G网络、蓝牙(Bluetooth)、Wi-Fi等无线或有线网络。需要指出的是,图8仅示出了具有部件20-23的计算机设备2,但是应理解的是,并不要求实施所有示出的部件,可以替代的实施更多或者更少的部件。
在本实施例中,存储于存储器21中的所述非接触性心率检测系统20还可以被分割为一个或者多个程序模块,所述一个或者多个程序模块被存储于存储器21中,并由一个或多个处理器(本实施例为处理器22)所执行,以完成本申请。
例如,图7示出了所述实现非接触性心率检测系统20实施例二的程序模块示意图,该实施例中,所述非接触性心率检测系统20可以被划分为获取模块200、提取模块202、选取模块204、放大模块206、分析模块208及计算模块210。其中,本申请所称的程序模块是指能够完成特定功能的一系列计算机程序指令段。所述程序模块200-210的具体功能在实施例二中已有详细描述,在此不再赘述。
实施例四
本实施例还提供一种非易失性计算机可读存储介质,如闪存、硬盘、多媒体卡、卡型存储器(例如,SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘、服务器、App应用商城等等,其上存储有计算机可读指令,程序被处理器执行时实现相应功能。本实施例的非易失性计算机可读存储介质用于存储非接触性心率检测系统20,被处理器执行时实现以下步骤:
获取待处理视频中的多帧人脸图像;
提取所述多帧人脸图像中的预设面部区域;
获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
根据所述目标频率计算得到心率值。
在一实施方式中,计算机可读指令被所述处理器执行时还实现以下步骤:
将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量;
选择能量最大的通道作为所述目标通道。
本申请实施例提供的非接触性心率检测方法、系统、设备及存储介质,通过从视频中选取人脸部分,并利用欧拉影像放大算法对像素值矩阵进行放大处理得到血管脉搏波序列,再分析血管脉搏波序列的波形,最后计算出心率值;本申请实施例可以将心率的计算与真实心率绝对值误差在3以内,提高了检测的精确性。
上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。
以上仅为本申请的优选实施例,并非因此限制本申请的专利范围,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种非接触性心率检测方法,包括:
    获取待处理视频中的多帧人脸图像;
    提取所述多帧人脸图像中的预设面部区域;
    获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
    通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
    分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
    根据所述目标频率计算得到心率值。
  2. 根据权利要求1所述的非接触性心率检测方法,获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道的步骤,包括:
    将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量;
    选择能量最大的通道作为所述目标通道。
  3. 根据权利要求1所述的非接触性心率检测方法,通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列的步骤,包括:
    将所述目标通道对应的像素值矩阵进行空间滤波,以得到不同的空间频率的基带,其中,采用低通滤波器进行空间滤波;
    根据高斯金字塔对所述基带进行平滑与下采样得到所述心血管脉搏波序列。
  4. 根据权利要求3所述的非接触性心率检测方法,根据高斯金字塔对所述基带进行平滑与下采样得到所述心血管脉搏波序列的步骤,包括:
    第一层高斯金字塔通过平滑与下采样获得二层高斯图像,高斯金字塔的截 至频率从上一层到下一层以因子2逐渐增加;
    直至第K-1层高斯金字塔通过平滑与下采样获得第K层高斯图像,得到所述心血管脉搏波时间序列。
  5. 根据权利要求1所述的非接触性心率检测方法,分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率的步骤,包括:
    选择频率带宽0.4~4Hz作为分析频段,将所述心血管脉搏波序列的频率波形进行带通滤波得到所述心血管脉搏波序列的频率波形的波峰,其中波峰峰值最大对应的频率为所述目标频率。
  6. 根据权利要求1所述的非接触性心率检测方法,获取待处理视频信息中的多帧人脸图像信息的步骤,包括:
    依据时间顺序逐帧获取所述视频信息中的每帧图像的人脸图像信息;
    统计有效图像的图像帧数,所述有效图像为所述视频信息中含人脸图像信息的一个或多个图像;
    当所述有效图像的图像帧数大于预设阈值时,则停止获取所述视频信息中的每帧图像的人脸图像信息。
  7. 一种非接触性心率检测系统,包括:
    获取模块,用于获取待处理视频中的多帧人脸图像;
    提取模块,用于提取所述多帧人脸图像中的预设面部区域;
    选取模块,用于获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
    放大模块,用于通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
    分析模块,用于分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
    计算模块,用于根据所述目标频率计算得到心率值。
  8. 根据权利要求7所述的非接触性心率检测系统,所述获取模块还用于:
    依据时间顺序逐帧获取所述视频信息中的每帧图像的人脸图像信息;
    统计有效图像的图像帧数,所述有效图像为所述视频信息中含人脸图像信息的一个或多个图像;
    当所述有效图像的图像帧数大于预设阈值时,则停止获取所述视频信息中的每帧图像的人脸图像信息。
  9. 根据权利要求7所述的非接触性心率检测系统,所述选取模块还用于:
    将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量;
    选择能量最大的通道作为所述目标通道。
  10. 根据权利要求7所述的非接触性心率检测系统,所述放大模块还用于:
    将所述目标通道对应的像素值矩阵进行空间滤波,以得到不同的空间频率的基带,其中,采用低通滤波器进行空间滤波;
    根据高斯金字塔对所述基带进行平滑与下采样得到所述心血管脉搏波序列。
  11. 根据权利要求10所述的非接触性心率检测系统,所述放大模块还用于:
    第一层高斯金字塔通过平滑与下采样获得二层高斯图像,高斯金字塔的截至频率从上一层到下一层以因子2逐渐增加;
    直至第K-1层高斯金字塔通过平滑与下采样获得第K层高斯图像,得到所述心血管脉搏波时间序列。
  12. 根据权利要求7所述的非接触性心率检测系统,所述分析模块还用于:
    选择频率带宽0.4~4Hz作为分析频段,将所述心血管脉搏波序列的频率波形进行带通滤波得到所述心血管脉搏波序列的频率波形的波峰,其中波峰峰值最大对应的频率为所述目标频率。
  13. 一种计算机设备,所述计算机设备包括存储器、处理器,所述存储器上存储有可在所述处理器上运行的计算机可读指令,所述计算机可读指令被所述处理器执行时实现以下步骤:
    获取待处理视频中的多帧人脸图像;
    提取所述多帧人脸图像中的预设面部区域;
    获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
    通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
    分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
    根据所述目标频率计算得到心率值。
  14. 根据权利要求13所述的计算机设备,所述计算机可读指令被所述处理器执行时实现以下步骤:
    依据时间顺序逐帧获取所述视频信息中的每帧图像的人脸图像信息;
    统计有效图像的图像帧数,所述有效图像为所述视频信息中含人脸图像信息的一个或多个图像;
    当所述有效图像的图像帧数大于预设阈值时,则停止获取所述视频信息中的每帧图像的人脸图像信息。
  15. 根据权利要求13所述的计算机设备,所述计算机可读指令被所述处理器执行时实现以下步骤:
    将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量;
    选择能量最大的通道作为所述目标通道。
  16. 根据权利要求13所述的计算机设备,所述计算机可读指令被所述处理器执行时实现以下步骤:
    将所述目标通道对应的像素值矩阵进行空间滤波,以得到不同的空间频率的基带,其中,采用低通滤波器进行空间滤波;
    根据高斯金字塔对所述基带进行平滑与下采样得到所述心血管脉搏波序列。
  17. 根据权利要求16所述的计算机设备,所述计算机可读指令被所述处理器执行时实现以下步骤:
    第一层高斯金字塔通过平滑与下采样获得二层高斯图像,高斯金字塔的截 至频率从上一层到下一层以因子2逐渐增加;
    直至第K-1层高斯金字塔通过平滑与下采样获得第K层高斯图像,得到所述心血管脉搏波时间序列。
  18. 根据权利要求13所述的计算机设备,所述计算机可读指令被所述处理器执行时实现以下步骤:
    选择频率带宽0.4~4Hz作为分析频段,将所述心血管脉搏波序列的频率波形进行带通滤波得到所述心血管脉搏波序列的频率波形的波峰,其中波峰峰值最大对应的频率为所述目标频率。
  19. 一种非易失性计算机可读存储介质,所述非易失性计算机可读存储介质内存储有计算机可读指令,所述计算机可读指令可被至少一个处理器所执行,以使所述至少一个处理器执行以下步骤:
    获取待处理视频中的多帧人脸图像;
    提取所述多帧人脸图像中的预设面部区域;
    获取所述预设面部区域的R通道、G通道和B通道的像素值矩阵,并从所述R通道、G通道和B通道中选取其中一个通道作为目标通道;
    通过欧拉影像放大对所述目标通道对应的像素值矩阵进行放大,以得到心血管脉搏波序列;
    分析所述心血管脉搏波序列的频率波形,选取所述频率波形的各个波峰中的最大波峰所对应的频率值作为目标频率;
    根据所述目标频率计算得到心率值。
  20. 根据权利要求19所述的非易失性计算机可读存储介质,所述计算机可读指令被所述处理器执行时还实现以下步骤:
    将所述R通道、G通道和B通道中各个通道的像素值矩阵通过快速傅里叶变换做频域变换以得到各个通道的通道能量;
    选择能量最大的通道作为所述目标通道。
PCT/CN2019/118082 2019-07-31 2019-11-13 非接触性心率检测方法、系统、设备及存储介质 Ceased WO2021017307A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910699269.6 2019-07-31
CN201910699269.6A CN110547783B (zh) 2019-07-31 2019-07-31 非接触性心率检测方法、系统、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2021017307A1 true WO2021017307A1 (zh) 2021-02-04

Family

ID=68736860

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/118082 Ceased WO2021017307A1 (zh) 2019-07-31 2019-11-13 非接触性心率检测方法、系统、设备及存储介质

Country Status (2)

Country Link
CN (1) CN110547783B (zh)
WO (1) WO2021017307A1 (zh)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112965060A (zh) * 2021-02-19 2021-06-15 加特兰微电子科技(上海)有限公司 生命特征参数的检测方法、装置和检测体征点的方法
CN113842128A (zh) * 2021-09-29 2021-12-28 北京清智图灵科技有限公司 一种基于多重滤波与混合放大的非接触心率检测装置
CN113989880A (zh) * 2021-10-18 2022-01-28 浙江大学 基于人脸视频的人体心率测量方法
CN114469036A (zh) * 2022-01-27 2022-05-13 无锡博奥玛雅医学科技有限公司 一种基于视频图像的远程心率监测方法及系统
CN114983415A (zh) * 2022-05-12 2022-09-02 华南理工大学 基于面部视频的工作记忆负荷水平识别方法
CN115205270A (zh) * 2022-07-25 2022-10-18 哈尔滨工业大学 基于图像处理的非接触式血氧饱和度检测方法及系统
CN115462800A (zh) * 2021-06-11 2022-12-13 广州视源电子科技股份有限公司 一种心电信号特征波形检测方法、装置、设备及存储介质
CN115919277A (zh) * 2022-12-06 2023-04-07 浙江大学 一种基于温度阈值的心率检测方法
CN116077062A (zh) * 2023-04-10 2023-05-09 中国科学院自动化研究所 心理状态感知方法及系统、可读存储介质
CN116758066A (zh) * 2023-08-14 2023-09-15 中国科学院长春光学精密机械与物理研究所 一种非接触测量心率方法、设备及介质
CN117158926A (zh) * 2023-09-21 2023-12-05 安徽大学 一种长距离非接触式的生理参数的检测方法、系统、装置
CN117694845A (zh) * 2024-02-06 2024-03-15 北京科技大学 基于融合特征增强的非接触式生理信号检测方法及装置

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111166313A (zh) * 2019-12-26 2020-05-19 中国电子科技集团公司电子科学研究院 一种心率测量方法、装置及可读存储介质
CN111510768B (zh) * 2020-04-26 2022-01-04 梁华智能科技(上海)有限公司 一种视频流的生命体征数据计算方法、设备及介质
TWI742696B (zh) * 2020-05-28 2021-10-11 國立臺灣科技大學 多參數生理訊號量測方法
CN114581358B (zh) * 2020-11-30 2025-07-25 华为技术有限公司 心率检测方法及电子设备
CN112580612B (zh) * 2021-02-22 2021-06-08 中国科学院自动化研究所 一种生理信号预测方法
CN113951816B (zh) * 2021-09-07 2024-04-12 广东省科学院健康医学研究所 基于光学视频信号分析的无创血管功能检测装置
CN113940632A (zh) * 2021-10-19 2022-01-18 展讯通信(天津)有限公司 健康指标检测方法和设备
CN114724068A (zh) * 2022-04-07 2022-07-08 中广核核电运营有限公司 人员状态监测方法、装置、设备、存储介质及程序产品
CN114943732B (zh) * 2022-07-01 2024-07-16 上海商汤临港智能科技有限公司 一种心率检测方法、装置、设备和存储介质
CN116019435B (zh) * 2022-12-27 2024-11-05 北京镁伽机器人科技有限公司 一种类心脏的心率确定方法、装置、电子设备及存储介质
CN119138868A (zh) * 2023-06-14 2024-12-17 华为技术有限公司 一种血压检测装置
CN116681700B (zh) * 2023-08-01 2023-10-31 首都医科大学附属北京天坛医院 用户心率和心率变异性的评估方法、装置和可读储存介质
CN119157504A (zh) * 2024-08-28 2024-12-20 中国人民解放军总医院 非接触式心率和呼吸率检测方法和装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130096439A1 (en) * 2011-10-14 2013-04-18 Industrial Technology Research Institute Method and system for contact-free heart rate measurement
CN107529646A (zh) * 2017-05-02 2018-01-02 广东工业大学 一种基于欧拉影像放大的无接触式心率测量方法及装置
CN109044322A (zh) * 2018-08-29 2018-12-21 北京航空航天大学 一种非接触式心率变异性测量方法
CN109259749A (zh) * 2018-08-29 2019-01-25 南京邮电大学 一种基于视觉摄像头的非接触式心率测量方法
CN109523545A (zh) * 2018-11-28 2019-03-26 荆门博谦信息科技有限公司 一种非接触式心率检测方法及系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110251493A1 (en) * 2010-03-22 2011-10-13 Massachusetts Institute Of Technology Method and system for measurement of physiological parameters
EP3495994A1 (en) * 2017-12-05 2019-06-12 Tata Consultancy Services Limited Face video based heart rate monitoring using pulse signal modelling and tracking
CN109009052A (zh) * 2018-07-02 2018-12-18 南京工程学院 基于视觉的嵌入式心率测量系统及其测量方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130096439A1 (en) * 2011-10-14 2013-04-18 Industrial Technology Research Institute Method and system for contact-free heart rate measurement
CN107529646A (zh) * 2017-05-02 2018-01-02 广东工业大学 一种基于欧拉影像放大的无接触式心率测量方法及装置
CN109044322A (zh) * 2018-08-29 2018-12-21 北京航空航天大学 一种非接触式心率变异性测量方法
CN109259749A (zh) * 2018-08-29 2019-01-25 南京邮电大学 一种基于视觉摄像头的非接触式心率测量方法
CN109523545A (zh) * 2018-11-28 2019-03-26 荆门博谦信息科技有限公司 一种非接触式心率检测方法及系统

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CAO JIANJIAN , FENG JUN , TANG WEN-MING , YU YING: "Color Space Selection of Non-contact Heart Rate Measurement", COMPUTER SCIENCE, vol. 44, no. 11A, 30 November 2017 (2017-11-30), pages 260 - 262+301, XP055776450 *
LIU, HONGCHENG ET AL: "Video-based Detection of Heart and Respiration Rates During Human Sleep", SCIENCE AND TECHNOLOGY & INNOVATION, no. 5, 31 March 2019 (2019-03-31), pages 1 - 4, XP009525737, DOI: 10.15913/j.cnki.kjycx.2019.05.001 *

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112965060A (zh) * 2021-02-19 2021-06-15 加特兰微电子科技(上海)有限公司 生命特征参数的检测方法、装置和检测体征点的方法
CN115462800A (zh) * 2021-06-11 2022-12-13 广州视源电子科技股份有限公司 一种心电信号特征波形检测方法、装置、设备及存储介质
CN113842128B (zh) * 2021-09-29 2023-09-26 北京清智图灵科技有限公司 一种基于多重滤波与混合放大的非接触心率检测装置
CN113842128A (zh) * 2021-09-29 2021-12-28 北京清智图灵科技有限公司 一种基于多重滤波与混合放大的非接触心率检测装置
CN113989880A (zh) * 2021-10-18 2022-01-28 浙江大学 基于人脸视频的人体心率测量方法
CN114469036A (zh) * 2022-01-27 2022-05-13 无锡博奥玛雅医学科技有限公司 一种基于视频图像的远程心率监测方法及系统
CN114983415A (zh) * 2022-05-12 2022-09-02 华南理工大学 基于面部视频的工作记忆负荷水平识别方法
CN115205270B (zh) * 2022-07-25 2023-10-24 哈尔滨工业大学 基于图像处理的非接触式血氧饱和度检测方法及系统
CN115205270A (zh) * 2022-07-25 2022-10-18 哈尔滨工业大学 基于图像处理的非接触式血氧饱和度检测方法及系统
CN115919277A (zh) * 2022-12-06 2023-04-07 浙江大学 一种基于温度阈值的心率检测方法
CN116077062A (zh) * 2023-04-10 2023-05-09 中国科学院自动化研究所 心理状态感知方法及系统、可读存储介质
CN116077062B (zh) * 2023-04-10 2023-06-30 中国科学院自动化研究所 心理状态感知方法及系统、可读存储介质
CN116758066A (zh) * 2023-08-14 2023-09-15 中国科学院长春光学精密机械与物理研究所 一种非接触测量心率方法、设备及介质
CN116758066B (zh) * 2023-08-14 2023-11-14 中国科学院长春光学精密机械与物理研究所 一种非接触测量心率方法、设备及介质
CN117158926A (zh) * 2023-09-21 2023-12-05 安徽大学 一种长距离非接触式的生理参数的检测方法、系统、装置
CN117694845A (zh) * 2024-02-06 2024-03-15 北京科技大学 基于融合特征增强的非接触式生理信号检测方法及装置
CN117694845B (zh) * 2024-02-06 2024-04-26 北京科技大学 基于融合特征增强的非接触式生理信号检测方法及装置

Also Published As

Publication number Publication date
CN110547783A (zh) 2019-12-10
CN110547783B (zh) 2022-05-17

Similar Documents

Publication Publication Date Title
WO2021017307A1 (zh) 非接触性心率检测方法、系统、设备及存储介质
WO2021169128A1 (zh) 眼底视网膜血管识别及量化方法、装置、设备及存储介质
CN111062963B (zh) 一种血管提取方法、系统、设备及存储介质
CN110751636B (zh) 一种基于改进型编解码网络的眼底图像视网膜动脉硬化检测方法
EP3453321A1 (en) Non-invasive method and system for estimating blood pressure from photoplethysmogram using statistical post-processing
Zou et al. An ultra-low power QRS complex detection algorithm based on down-sampling wavelet transform
CN112634231B (zh) 一种图像分类方法、装置、终端设备和存储介质
CN107292835B (zh) 一种眼底图像视网膜血管自动矢量化的方法及装置
US12561811B2 (en) Method and apparatus of nidus segmentation, electronic device, and storage medium
CN111358455A (zh) 一种多数据源的血压预测方法和装置
CN106491114B (zh) 一种心率检测方法及装置
CN118121176B (zh) 一种用于无创心排监测的数据分析方法及系统
CN116196013B (zh) 心电数据处理方法、装置、计算机设备与存储介质
CN111915515A (zh) 超声图像中噪声去除的方法、超声设备及存储介质
CN115147360B (zh) 一种斑块分割方法、装置、电子设备及可读存储介质
CN117357080B (zh) 近红外光谱信号去噪方法及装置、终端设备、存储介质
CN110916649B (zh) 一种长程心电散点图的处理装置、处理方法及检测装置
CN110874597B (zh) 一种眼底图像血管特征提取方法、设备、系统和存储介质
WO2017096597A1 (zh) 心电信号处理方法及装置
CN115067967B (zh) 心拍信号基准点确定方法、心拍类型识别方法及装置
CN118592919B (zh) 基于监护仪屏幕波形摄像的医疗辅助监测方法及装置
JP3647970B2 (ja) 領域抽出装置
CN116823752B (zh) 基于力学参量的脑网络构建方法、系统、介质及设备
CN111062943B (zh) 斑块稳定性的确定方法、装置及医疗设备
CN118229555B (zh) 一种图像融合方法、装置、设备及计算机可读存储介质

Legal Events

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

Ref document number: 19939823

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19939823

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 08.08.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 19939823

Country of ref document: EP

Kind code of ref document: A1