A heart rate detection method, device, apparatus and storage medium

Abstract

Embodiments of the present disclosure provide a heart rate detection method, device and equipment, and a storage medium, wherein the method comprises: acquiring a target image sequence, the target image sequence containing multiple images of a target object; extracting an initial signal representing heartbeat fluctuation of the target object based on the target image sequence; screening multiple signal segments with interval values of wave intervals meeting interval value conditions from signal segments of the initial signal, splicing the multiple signal segments to obtain a heartbeat fluctuation signal, the signal segment being a signal between two adjacent wave crests or a signal between two adjacent wave troughs in the initial signal; and obtaining a heart rate of the target object according to the heartbeat fluctuation signal. The method can accurately predict the heart rate of the target object through the image of the target object.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, and specifically to a heart rate detection method, apparatus, device, and storage medium. Background Technology

[0002] Heart rate is the most direct indicator of health. Traditional heart rate detection mostly involves wearing a heart rate monitoring device and using methods such as blood oxygenation or photoplethysmography, which requires contact with the skin. The detection process is inconvenient and expensive.

[0003] Currently, most vision-based non-contact heart rate detection methods use the rPPG (remote photoplethysmography) method. This method uses a camera to capture the subtle changes in skin brightness caused by blood flow during heartbeats, thereby detecting periodic heartbeats. However, the detected heart rate information is easily affected by facial movements or changes in lighting, especially in scenarios with frequent lighting changes, such as outdoors or inside a car, leading to inaccurate heart rate information. Summary of the Invention

[0004] In view of this, embodiments of the present disclosure provide at least one heart rate detection method, apparatus, device, and storage medium.

[0005] Specifically, the embodiments of this disclosure are implemented through the following technical solutions:

[0006] Firstly, a heart rate detection method is provided, the method comprising:

[0007] Obtain a target image sequence, wherein the target image sequence contains multiple frames of images of the target object;

[0008] Based on the target image sequence, an initial signal characterizing the heartbeat fluctuation of the target object is extracted;

[0009] From the signal segments of the initial signal, multiple signal segments whose wave spacing values ​​meet the spacing value conditions are selected, and the multiple signal segments are spliced ​​together to obtain the heartbeat fluctuation signal. The signal segment is the signal between two adjacent wave peaks or the signal between two adjacent wave troughs in the initial signal.

[0010] The heart rate of the target object is obtained based on the heartbeat fluctuation signal.

[0011] In some optional embodiments, before selecting multiple signal segments from the signal segments of the initial signal whose interval values ​​meet the interval value condition, and splicing the multiple signal segments to obtain the heartbeat fluctuation signal, the method further includes:

[0012] Determine the signal between every two adjacent peaks and troughs in the initial signal, and the signal between every two adjacent troughs and peaks;

[0013] The amplitudes of the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are scaled to the same range to obtain an initial signal with uniform amplitude.

[0014] In some optional embodiments, the step of filtering multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition includes:

[0015] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0016] Subtract a preset error value from the average spacing to determine the minimum threshold for wave spacing; add the preset error value to the average spacing to determine the maximum threshold for wave spacing.

[0017] From the signal segments of the initial signal, select multiple signal segments whose inter-wave spacing values ​​reach the minimum inter-wave spacing threshold and do not exceed the maximum inter-wave spacing threshold.

[0018] In some optional embodiments, the step of filtering multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition includes:

[0019] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0020] Based on a preset scaling factor and the number of wave spacings in the initial signal, the number of signal segments to be deleted in the initial signal is determined as a first number.

[0021] In order of increasing inter-wave spacing, a second number of signal segments are deleted from at least one signal segment in the initial signal whose inter-wave spacing value is less than the average inter-wave spacing value.

[0022] According to the order of wave spacing from largest to smallest, a third number of signal segments are deleted from at least one signal segment in the initial signal whose wave spacing value is greater than the average spacing value; wherein, the sum of the second number and the third number is the first number;

[0023] The remaining signal segments are identified as the multiple signal segments obtained through filtering.

[0024] In some optional embodiments, before obtaining the heart rate of the target object based on the heart rate fluctuation signal, the method further includes:

[0025] Based on the position of the peak or trough in the heartbeat fluctuation signal, the heartbeat fluctuation signal is divided into multiple first signals, each of which contains at least one cycle of signal.

[0026] The mean value of the first spacing corresponding to each segment of the first signal is calculated based on the spacing between each two adjacent peaks or troughs in each segment of the first signal.

[0027] The average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal.

[0028] Select the first signal whose first spacing mean is closest to the second spacing mean from each segment of the first signal as the reference signal;

[0029] For each segment of the first signal other than the reference signal, calculate the similarity between the first signal and the reference signal;

[0030] By splicing the reference signal and at least one segment of the first signal whose similarity to the reference signal meets a preset condition, a new heartbeat fluctuation signal is obtained.

[0031] In some alternative embodiments, each segment of the first signal has the same number of peaks and the same number of troughs.

[0032] In some optional embodiments, the images in the target image sequence contain regions of interest for a face, which include at least one of the following regions: a forehead region, a left cheek region, and a right cheek region;

[0033] The acquisition of the target image sequence includes:

[0034] Based on the video stream obtained from the target object during the first time period, a target image sequence consisting of multiple video frames containing the target object is obtained;

[0035] The step of extracting the initial signal characterizing the heartbeat fluctuation of the target object based on the target image sequence includes:

[0036] For each video frame in the target image sequence, the following processing is performed: based on face detection of the video frame, a face detection box of the target object is obtained; based on face region detection of the face region in the face detection box, face key points of the target object are determined; based on the face key points, the region of interest of the face contained in the video frame is determined; the sum of the average pixel values ​​of each region in the region of interest of the face is calculated as the signal value of the video frame.

[0037] A timing signal is obtained based on multiple sequentially consecutive signal values ​​within the first time period;

[0038] The blood volume pulse (rPPG) signal is extracted from the time-series signal and used as the initial signal.

[0039] In some optional embodiments, obtaining the heart rate of the target object based on the heartbeat fluctuation signal includes:

[0040] The heart rate of the target object is selected from the multiple frequencies by performing a Fourier transform on the heart rate fluctuation signal.

[0041] Secondly, a heart rate detection device is provided, the device comprising:

[0042] The image acquisition module is used to acquire a target image sequence, wherein the target image sequence contains multiple frames of images of the target object;

[0043] The signal extraction module is used to extract an initial signal representing the heartbeat fluctuation of the target object based on the target image sequence;

[0044] The signal filtering module is used to filter multiple signal segments from the signal segments of the initial signal whose interval values ​​meet the interval value conditions, and splice the multiple signal segments to obtain the heartbeat fluctuation signal. The signal segment is the signal between two adjacent peaks or the signal between two adjacent troughs in the initial signal.

[0045] The heart rate determination module is used to obtain the heart rate of the target object based on the heartbeat fluctuation signal.

[0046] In some optional embodiments, before selecting multiple signal segments from the signal segments of the initial signal whose interval values ​​meet the interval value conditions, and splicing the multiple signal segments to obtain the heartbeat fluctuation signal, the signal filtering module is further configured to determine the signal between every two adjacent peaks and troughs in the initial signal, and the signal between every two adjacent troughs and peaks.

[0047] The amplitudes of the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are scaled to the same range to obtain an initial signal with uniform amplitude.

[0048] In some optional embodiments, when the signal filtering module is used to filter multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition, it is specifically used to:

[0049] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0050] Subtract a preset error value from the average spacing to determine the minimum threshold for wave spacing; add the preset error value to the average spacing to determine the maximum threshold for wave spacing.

[0051] From the signal segments of the initial signal, select multiple signal segments whose inter-wave spacing values ​​reach the minimum inter-wave spacing threshold and do not exceed the maximum inter-wave spacing threshold.

[0052] In some optional embodiments, when the signal filtering module is used to filter multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition, it is specifically used to:

[0053] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0054] Based on a preset scaling factor and the number of wave spacings in the initial signal, the number of signal segments to be deleted in the initial signal is determined as a first number.

[0055] In order of increasing inter-wave spacing, a second number of signal segments are deleted from at least one signal segment in the initial signal whose inter-wave spacing value is less than the average inter-wave spacing value.

[0056] According to the order of wave spacing from largest to smallest, a third number of signal segments are deleted from at least one signal segment in the initial signal whose wave spacing value is greater than the average spacing value; wherein, the sum of the second number and the third number is the first number;

[0057] The remaining signal segments are identified as the multiple signal segments obtained through filtering.

[0058] In some optional embodiments, before obtaining the heart rate of the target object based on the heartbeat fluctuation signal, the signal filtering module is further configured to divide the heartbeat fluctuation signal into multiple first signals based on the position of the peak or trough in the heartbeat fluctuation signal, each first signal containing at least one cycle of signal;

[0059] The mean value of the first spacing corresponding to each segment of the first signal is calculated based on the spacing between each two adjacent peaks or troughs in each segment of the first signal.

[0060] The average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal.

[0061] Select the first signal whose first spacing mean is closest to the second spacing mean from each segment of the first signal as the reference signal;

[0062] For each segment of the first signal other than the reference signal, calculate the similarity between the first signal and the reference signal;

[0063] By splicing the reference signal and at least one segment of the first signal whose similarity to the reference signal meets a preset condition, a new heartbeat fluctuation signal is obtained.

[0064] In some alternative embodiments, each segment of the first signal has the same number of peaks and the same number of troughs.

[0065] In some optional embodiments, the images in the target image sequence contain regions of interest for a face, which include at least one of the following regions: a forehead region, a left cheek region, and a right cheek region;

[0066] The image acquisition module is specifically used for:

[0067] Based on the video stream obtained from the target object during the first time period, a target image sequence consisting of multiple video frames containing the target object is obtained;

[0068] The signal extraction module is specifically used for:

[0069] For each video frame in the target image sequence, the following processing is performed: based on face detection of the video frame, a face detection box of the target object is obtained; based on face region detection of the face region in the face detection box, face key points of the target object are determined; based on the face key points, the region of interest of the face contained in the video frame is determined; the sum of the average pixel values ​​of each region in the region of interest of the face is calculated as the signal value of the video frame.

[0070] A timing signal is obtained based on multiple sequentially consecutive signal values ​​within the first time period;

[0071] The blood volume pulse (rPPG) signal is extracted from the time-series signal and used as the initial signal.

[0072] In some optional embodiments, the heart rate determination module is specifically used for:

[0073] The heart rate of the target object is selected from the multiple frequencies by performing a Fourier transform on the heart rate fluctuation signal.

[0074] Thirdly, an electronic device is provided, the device including a memory and a processor, the memory being used to store computer instructions executable on the processor, and the processor being used to implement the heart rate detection method according to any embodiment of the present disclosure when executing the computer instructions.

[0075] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the heart rate detection method according to any embodiment of the present disclosure.

[0076] The heart rate detection method provided in this embodiment obtains the initial fluctuation signal by periodically changing the skin pixel values ​​of the region of interest on the face of the target object, and then filters it according to the wave interval condition to obtain the heart rate fluctuation signal, thereby obtaining the heart rate of the target object. It can remove unstable signals caused by interference factors such as changes in lighting or facial movement, and retain stable signals, thereby accurately predicting the heart rate of the target object during that period of time. Attached Figure Description

[0077] To more clearly illustrate the technical solutions in one or more embodiments or related technologies of this disclosure, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in one or more embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0078] Figure 1 This is a flowchart illustrating a heart rate detection method according to at least one embodiment of the present disclosure;

[0079] Figure 2 This is a flowchart illustrating another heart rate detection method according to at least one embodiment of this disclosure;

[0080] Figure 3 This is a flowchart illustrating a signal filtering method according to at least one embodiment of the present disclosure;

[0081] Figure 4 This is a block diagram illustrating a heart rate detection device according to at least one embodiment of the present disclosure;

[0082] Figure 5 This is a schematic diagram of the hardware structure of an electronic device shown in at least one embodiment of the present disclosure. Detailed Implementation

[0083] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this specification as detailed in the appended claims.

[0084] The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of this specification. The singular forms “a,” “the,” and “the” as used in this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0085] It should be understood that although the terms first, second, third, etc., may be used in this specification to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this specification, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0086] Current remote heart rate detection methods mostly use the rPPG method, which uses a camera to capture subtle changes in skin brightness to detect periodic heartbeats. These changes in skin brightness are caused by blood flow due to the heartbeat. When a light source illuminates the tissue, a photodetector captures the minute changes in the intensity of light reflected or transmitted from the blood flow, generating the so-called rPPG signal. Light absorption follows Beer-Lambert's law, which states that the amount of light absorbed by the blood is proportional to the amount of light penetrating the skin and the concentration of hemoglobin in the blood. During the heartbeat cycle, minute changes in hemoglobin concentration cause fluctuations in the amount of light absorbed by the blood vessels, leading to changes in the intensity of light reflected from the skin. However, heart rate information from the face is affected by facial movement, changes in lighting, and other environmental factors, especially in scenarios with frequent lighting changes, such as outdoors or inside a car. This makes the extracted rPPG signal inaccurate, resulting in inaccurate heart rate detection. The heart rate detection method proposed in this disclosure can process the extracted signal to obtain an accurate heart rate.

[0087] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a heart rate detection method according to at least one embodiment of the present disclosure, the method comprising the following steps:

[0088] In step 102, a target image sequence is obtained, which contains multiple frames of images of the target object.

[0089] The target image sequence comprises multiple frames of images of the target object within a first time period. The target image sequence can be images of the same target object at various time points within a video stream acquired during the first time period. This embodiment does not limit the acquisition method of the target image sequence. The acquisition time of the multiple frames is continuous; for example, the first time period can be 30 seconds. For a video stream containing the target object for 30 seconds, at a frame rate of 30 frames per second, the resulting target image sequence contains 900 frames.

[0090] The images in the target image sequence can be images containing any skin region of the target object, such as the neck or hands.

[0091] In one embodiment, the images in the target image sequence include regions of interest (ROIs) for the face of the target object. The ROI can be any region of facial skin, and the pixel values ​​of the skin can be grayscale values, brightness values, or intensity values, etc. For example, the forehead and cheek regions, which have more rPPG information, can be selected as ROIs.

[0092] The heart rate detection method in this embodiment is applicable to various scenarios, such as hospitals, home sleep monitoring, and heart rate detection in vehicles. The target object can be in a moving or stationary state, indoors or outdoors. Image acquisition is performed on the target object with its consent.

[0093] In step 104, an initial signal characterizing the heartbeat fluctuation of the target object is extracted based on the target image sequence.

[0094] The initial signal is a time-varying signal composed of multiple cycles. Each cycle is not identical in signal strength or length. The initial signal contains information about the periodic changes in skin pixel values ​​caused by blood flow from the heartbeat, and may also include interference from environmental factors such as changes in lighting and facial movement. For example, when the target is inside a car, the car vibrates during movement, causing the target's face to vibrate accordingly, and the lighting changes drastically as the vehicle moves.

[0095] In this embodiment, the initial signal can be an rPPG signal extracted from the target image sequence using any method. For example, it can be an rPPG signal extracted using neural networks such as 3D CNN (Convolutional Neural Network) or 2D CNN, or an rPPG signal extracted using traditional methods such as blind source separation, PBV (model-based method), or 2SR (Spatial subspace rotation). The initial signal can also be a signal containing heart rate information obtained by using other types of methods based on the periodic changes of skin pixel values ​​in the region of interest of the face.

[0096] In step 106, multiple signal segments whose wave spacing values ​​meet the spacing value conditions are selected from the signal segments of the initial signal, and the multiple signal segments are spliced ​​together to obtain the heartbeat fluctuation signal.

[0097] The signal segment is the signal between two adjacent peaks or two adjacent troughs in the initial signal, and each signal segment represents one cycle of the signal. The inter-wave spacing is the distance between two adjacent cycles of the signal in the initial signal, or the distance between two adjacent signal segments. It can be represented by the distance between the peaks or troughs of the signals in two adjacent cycles, or by the distance between other points at the same position on the wave of the signals in two adjacent cycles.

[0098] This embodiment does not impose specific spacing value conditions. For example, the spacing value range of the wave spacing can be set according to the experience of those skilled in the art, and the signals corresponding to the wave spacing values ​​that exceed the range in the initial signal can be deleted. Alternatively, a scaling factor can be set according to the experience of those skilled in the art. This scaling factor is used to calculate the number of signals corresponding to the wave spacing to be deleted. For example, if there are 100 wave spacings in the initial wave signal, when the scaling factor is 10%, the number of signals corresponding to the wave spacing to be deleted is 10. By sorting all the spacing values, 5 can be deleted starting from the smallest spacing value and 5 can be deleted starting from the largest spacing value.

[0099] The signals corresponding to the remaining interval values ​​after deletion, i.e., the signals with at least one cycle obtained through filtering, are then spliced ​​together to form the heartbeat fluctuation signal. It should be noted that when deleting the signals corresponding to the interval values, one can delete the signal between two adjacent peaks; in this case, during splicing, the peaks at both ends of the signals from different cycles are simply joined. Similarly, one can delete the signal between two adjacent troughs; in this case, the troughs at both ends of the signals from different cycles are simply joined.

[0100] Interference from factors such as facial movement and changes in lighting can cause variations in the interval value of the wave interval. Therefore, by filtering the interval value of the wave interval, unstable signals caused by interference factors can be filtered out, thereby obtaining a more stable heart rate signal.

[0101] In step 108, the heart rate of the target object is obtained based on the heartbeat fluctuation signal.

[0102] The heart rate fluctuation signal contains the signal of periodic changes in skin pixel values ​​caused by blood flow generated by the heartbeat. By performing a Fourier transform on this signal, the heart rate of the target object in the first time period can be obtained.

[0103] In one example, since there may be interference from other factors in the heartbeat fluctuation signal, after performing a Fourier transform on the signal, multiple frequencies may be obtained. Therefore, after performing a Fourier transform on the heartbeat fluctuation signal, the frequency with the largest power spectrum can be selected as the heart rate of the target object from the multiple frequencies obtained.

[0104] In other examples, the heart rate fluctuation signal can be further processed by bandpass filtering, such as Butterworth filtering, which limits the signal in the heart rate frequency band (usually 0.4Hz-4Hz) to pass through the filter. The signal in the heart rate frequency band of the heart rate fluctuation signal can then be obtained by performing a Fourier transform on the signal.

[0105] The heart rate detection method provided in this embodiment obtains the initial signal from the periodic changes of skin pixel values ​​in the region of interest of the target object's face, and then filters it according to the spacing value condition to obtain the heart rate fluctuation signal, thereby obtaining the heart rate of the target object. It can remove unstable signals caused by changes in the spacing value of the wave interval due to interference factors such as changes in lighting or facial movement, and retain stable signals, thereby accurately predicting the heart rate of the target object during that period.

[0106] Figure 2 This is a flowchart illustrating at least one embodiment of a heart rate detection method according to the present disclosure, which describes the heart rate detection process in more detail. Figure 2 As shown, the method may include the following processing. It should be noted that this embodiment does not limit the execution order of each step.

[0107] In step 202, a target image sequence consisting of multiple video frames containing the target object is obtained based on the video stream acquired from the target object during the first time period.

[0108] For example, when the target object needs to have its heart rate detected, the camera that captures images of the target object is turned on to obtain a video stream, thereby obtaining a target image sequence consisting of multiple video frames containing the target object within the first time period.

[0109] In step 204, each video frame in the target image sequence is processed to obtain the initial signal.

[0110] In one implementation, each video frame in the target image sequence may be processed as follows:

[0111] Based on face detection of the video frame, a face detection bounding box of the target object is obtained; based on face region detection of the face region in the face detection bounding box, face key points of the target object are determined; based on the face key points, the region of interest of the face contained in the video frame is determined; the sum of the average pixel values ​​of each region in the region of interest of the face is calculated as the signal value of the video frame.

[0112] In practice, face detection algorithms, such as the face detector in OpenCV (an open-source computer vision library), the face detector in dlib (a face recognition library), or other face detection algorithms, are used to detect faces in each frame of the video stream to obtain the target object's face bounding box. Facial landmark detection algorithms are then used to detect facial landmarks within the face bounding box; for example, a 68-point landmark detection algorithm can be used to obtain multiple facial landmarks of the target object. These landmarks are used to accurately locate different regions of the face. In other examples, face detection can be skipped, and facial landmark detection algorithms can be used directly to detect facial landmarks in the video frames; alternatively, face region detection can be used directly instead of face detection and facial landmark detection to obtain the image of the region of interest (ROI) of the face.

[0113] The face image is an image containing a region of interest (ROI) for the face. The ROI includes at least one of the following regions: the forehead region, the left cheek region, and the right cheek region. In this embodiment, the selected ROI includes the above three regions. In other examples, only one or two regions may be selected, or other regions of the face may be selected as the ROI.

[0114] Based on the forehead, left cheek, and right cheek key points in the facial landmarks, the forehead, left cheek, and right cheek regions of the face in the video frame can be located and identified.

[0115] The average pixel value of each region within the region of interest (ROI) of the face is calculated. The average pixel values ​​of the images within the forehead, left cheek, and right cheek regions are calculated separately and summed to obtain the signal value of that video frame. When there is only one ROI, summation is unnecessary, and the average pixel value of that region's image is directly used as the signal value of the video frame. In other examples, other methods for calculating the signal value can be used, such as a weighted average of the average pixel values ​​of the images from each region. This embodiment does not impose any limitations on this approach.

[0116] Thus, the above processing is performed on each video frame in the target image sequence to obtain its corresponding signal value.

[0117] In other embodiments, when the target object is stationary, or when the position of the target object's facial region of interest in the video frame of the target image sequence does not change, the following processing can be performed on only one video frame in the target image sequence: based on face detection of the video frame, a face detection box of the target object is obtained; based on facial key point detection of the face region in the face detection box, facial key points of the target object are determined; and based on the facial key points, the position of the facial region of interest in the video frame is determined.

[0118] Based on the fixed position of the region of interest (ROI) of the face in the video frame, the image of the ROI of the face in each video frame of the target image sequence is determined, and then the signal value corresponding to each video frame is calculated.

[0119] After obtaining the signal value corresponding to each video frame in the target image sequence, a time-series signal is obtained based on multiple consecutive signal values ​​within the first time period. For example, for a video stream with a frame rate of 30 frames per second within the first time period (30s), the signal length of the time-series signal composed of the signal values ​​corresponding to each frame is 30*30=900.

[0120] Extract the rPPG signal from the time-series signal as the initial signal. The extraction method is not limited to the time-series signal; it can be performed using a neural network or traditional rPPG signal extraction methods, such as chrom, ICA, and PCA algorithms, to obtain the initial signal.

[0121] In step 206, the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are determined in the initial signal; the amplitudes of the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are scaled to the same value range to obtain an initial signal with uniform amplitude.

[0122] Due to various reasons such as interference or errors during signal extraction, the amplitudes of the signals in different cycles of the initial signal are not consistent. If the signals are directly selected and stitched together based on the spacing condition, seamless stitching cannot be achieved because the amplitudes of peaks, troughs, and other points on the waves are inconsistent, affecting the accuracy of the calculated heart rate. Therefore, this step unifies the amplitudes of the signals in each cycle, adjusting the amplitudes of all peaks and troughs in the initial signal to be consistent, to facilitate subsequent processing.

[0123] For example, calculate all the maxima and minima (corresponding to the peaks and troughs of the signal) on the initial signal, and obtain the signal between each pair of adjacent maxima and minima, and between each pair of adjacent minima and maxima (equivalent to half a cycle of signal) based on each pair of adjacent maxima and minima. Then, scale the above signals from their original value range to a fixed value range, for example, scale them all to the range of 0-25 to obtain an initial signal with the same amplitude. In this initial signal with unified amplitude, the amplitude of each peak is 25 and the amplitude of each trough is 0.

[0124] This processing of the initial signal can improve the signal-to-noise ratio, and the heart rate calculated from the processed initial signal will be more accurate.

[0125] In step 208, multiple signal segments whose wave spacing values ​​meet the spacing value conditions are selected from the signal segments of the initial signal, and the multiple signal segments are spliced ​​together to obtain the heartbeat fluctuation signal.

[0126] In one example, the filtering based on the spacing value condition can be as follows: calculate the average spacing based on the spacing between every two adjacent peaks in the initial signal, or calculate the average spacing based on the spacing between every two adjacent troughs in the initial signal; subtract a preset error value from the average spacing to determine the minimum spacing threshold, add the preset error value to the average spacing to determine the maximum spacing threshold; and filter out multiple signal segments from the signal segments of the initial signal whose spacing values ​​reach the minimum spacing threshold and do not exceed the maximum spacing threshold.

[0127] For example, the spacing between any two adjacent maxima can be calculated to obtain multiple spacing values, or the spacing between any two adjacent minima can be calculated to obtain multiple spacing values. The average spacing value is then calculated based on these multiple spacing values. Filtering can be performed based on the absolute error between the spacing values ​​and the average spacing value. A threshold for absolute error is preset. The maximum spacing threshold is obtained by adding the threshold to the average spacing value, and the minimum spacing threshold is obtained by subtracting the threshold from the average spacing value. Signals whose spacing reaches the minimum spacing threshold but does not exceed the maximum spacing threshold are then filtered out. For example, when the absolute error threshold is 5, signals corresponding to spacings with an absolute error of more than 5 relative to the average spacing value can be deleted, and the remaining signals are retained.

[0128] In one example, the filtering based on the spacing value condition can be as follows: calculate the average spacing based on the spacing between every two adjacent peaks in the initial signal, or calculate the average spacing based on the spacing between every two adjacent troughs in the initial signal; determine the number of signal segments to be deleted in the initial signal as a first number based on a preset proportional coefficient and the number of spacings in the initial signal; delete a second number of signal segments from at least one signal segment in the initial signal whose spacing value is less than the average spacing value, in ascending order of spacing; delete a third number of signal segments from at least one signal segment in the initial signal whose spacing value is greater than the average spacing value, in descending order of spacing; wherein the sum of the second number and the third number is the first number; and determine the remaining signal segments as the multiple signal segments obtained through filtering.

[0129] In this example, the spacing between adjacent maxima (peaks) can be used to calculate the spacing between wave intervals. In other examples, the spacing between adjacent minima (troughs) can be used. Following the method in the previous example, the average spacing is calculated, for example, an average spacing of 50. The preset scaling factor can be set by those skilled in the art according to actual needs, for example, it can be set to 30%. The number of wave intervals in the initial signal is generally the same as the number of signal periods, or differs by one. This can be obtained by calculating the total number of wave intervals between adjacent peaks, or by directly calculating the total number of peaks. Assuming that the initial wave signal in this example contains a total of 100 intervals between adjacent maxima, that is, the number of wave intervals is 100, meaning the initial wave signal contains 100 signal segments. The signal segment corresponding to one wave interval, that is, one period of signal, is referred to here as a signal segment for ease of description.

[0130] Multiply the total number of wave spacings by the scaling factor to obtain the first number of signal segments to be deleted. For example, multiplying the number of wave spacings of 100 by the scaling factor of 30% gives the first number of 30.

[0131] Sort the intervals less than and greater than the average interval of 50. When both the second and third numbers are 15, remove 15 intervals that deviate most from 50 from both the intervals less than and greater than 50. That is, remove the top 15 intervals with the largest deviations from 50 and the top 15 intervals with the smallest deviations, leaving 70 segments of signal composed of adjacent maxima. The second and third numbers can be set arbitrarily. For example, if there are 4 intervals less than 50, the second number can be 4 and the third number can be 26, and their sum is the first number.

[0132] In one implementation, after processing the initial signal according to the interval value condition to obtain the heartbeat fluctuation signal, the following processing can be used for further filtering to further remove unstable signals, such as... Figure 3 As shown, it includes the following steps:

[0133] In step 302, based on the position of the peak or trough in the heartbeat fluctuation signal, the heartbeat fluctuation signal is divided into multiple first signals with the same number of periods.

[0134] In this embodiment, each first signal segment contains at least one period of signal. Each first signal segment has the same number of peaks and troughs to ensure more reliable subsequent similarity comparisons. In other examples, each first signal segment may contain different numbers of peaks and the same number of troughs, or the same number of peaks and different numbers of troughs.

[0135] For example, based on the location of the peaks, the heartbeat fluctuation signal can be divided into 10 segments of the first signal, each with the same number of peaks and troughs, with each segment beginning and ending at a peak. Alternatively, based on the location of the troughs, the heartbeat fluctuation signal can be divided into 10 segments of the first signal, each with the same number of peaks and troughs.

[0136] In step 304, the average first spacing value corresponding to each segment of the first signal is calculated based on the wave spacing between each two adjacent peaks or troughs in each segment of the first signal.

[0137] When segmenting according to the location of the peak, this step can calculate the average spacing between two adjacent peaks in each segment of the first signal, which serves as the average first spacing for each segment. When segmenting according to the location of the trough, this step can calculate the average spacing between two adjacent troughs in each segment of the first signal, which serves as the average first spacing for each segment.

[0138] In step 306, the average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal.

[0139] Calculate the average distance between two adjacent peaks or troughs in the heartbeat fluctuation signal obtained after preprocessing, i.e., the second average distance u.

[0140] In step 308, the first spacing average value that is closest to the second spacing average value is selected from the first spacing average values ​​corresponding to each segment of the first signal, and the first signal corresponding to the first spacing average value is used as the reference signal.

[0141] The signal segment with the closest average spacing to u from multiple first signal segments is selected as the reference signal.

[0142] In step 310, for each segment of the first signal other than the reference signal, the similarity between the first signal and the reference signal is calculated.

[0143] The similarity between the first signal in each of the 10 bands other than the reference signal and the reference signal can be calculated. For example, the similarity between the first signal and the reference signal can be obtained by calculating the l2 distance, Euclidean distance and cosine similarity between the signals.

[0144] In step 312, the reference signal and at least one segment of the first signal that meets the similarity condition are spliced ​​together to obtain a new heartbeat fluctuation signal.

[0145] One approach is to select the five most similar first signal segments and a reference signal, and then splice them together to form the final heartbeat fluctuation signal. By calculating the similarity, stable signals can be better selected.

[0146] In step 210, the heartbeat fluctuation signal is subjected to Fourier transform, and the frequency with the largest power spectrum is selected from the multiple frequencies obtained as the heart rate of the target object.

[0147] The heart rate detection method provided in this disclosure calculates the signal value of a video frame by detecting the pixel values ​​of the region of interest of the target object's face. The initial signal is extracted from the signal value and filtered according to the spacing value condition to remove unstable signals and obtain a heart rate fluctuation signal. The unstable signals can be further filtered by similarity to obtain a heart rate fluctuation signal composed of stable signals. This makes the calculated heart rate of the target object more accurate. It can remove unstable signals caused by changes in the wave spacing due to interference factors such as changes in lighting or facial movement, and retain stable signals to accurately predict the heart rate of the target object during that period.

[0148] In one embodiment, the heart rate detection method of this embodiment can be used in a driver detection system or a passenger detection system to perform contactless heart rate detection on people in the vehicle cabin. In the above embodiment, when obtaining a target image sequence composed of multiple video frames containing the target object based on the video stream acquired from the target object within a first time period, it can be: obtaining a target image sequence composed of multiple video frames containing the target object based on the video stream acquired from the target object within the vehicle cabin within the first time period; the target image includes: images of the people in the vehicle cabin. For example, when people in the vehicle cabin have a need for heart rate detection, with their consent, images of the people in the vehicle cabin can be acquired in real time through a camera in the vehicle cabin, and heart rate detection can be performed based on the acquired images. For people in the vehicle cabin, the acquired images are greatly affected by changes in lighting, and because the vehicle will vibrate during operation, this method can effectively filter out the influence of the above-mentioned lighting and vibration, thereby accurately performing heart rate detection based on time-series continuous images.

[0149] Those skilled in the art will understand that, in the above-described method of the specific implementation, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0150] like Figure 4 As shown, Figure 4 This is a block diagram illustrating a heart rate detection device according to at least one embodiment of the present disclosure, the device comprising:

[0151] Image acquisition module 41 is used to acquire a target image sequence, wherein the target image sequence contains multiple frames of images of the target object;

[0152] The signal extraction module 42 is used to extract an initial signal representing the heartbeat fluctuation of the target object based on the target image sequence;

[0153] Signal filtering module 43 is used to filter multiple signal segments from the signal segments of the initial signal whose wave spacing values ​​meet the spacing value conditions, and splice the multiple signal segments to obtain a heartbeat fluctuation signal. The signal segment is the signal between two adjacent wave peaks or the signal between two adjacent wave troughs in the initial signal.

[0154] The heart rate determination module 44 is used to obtain the heart rate of the target object based on the heartbeat fluctuation signal.

[0155] In some optional embodiments, before selecting multiple signal segments whose interval values ​​meet the interval value conditions from the signal segments of the initial signal and splicing the multiple signal segments to obtain the heartbeat fluctuation signal, the signal filtering module 43 is further used to determine the signal between every two adjacent peaks and troughs in the initial signal, and the signal between every two adjacent troughs and peaks.

[0156] The amplitudes of the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are scaled to the same range to obtain an initial signal with uniform amplitude.

[0157] In some optional embodiments, when the signal filtering module 43 filters multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition, it is specifically used for:

[0158] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0159] Subtract a preset error value from the average spacing to determine the minimum threshold for wave spacing; add the preset error value to the average spacing to determine the maximum threshold for wave spacing.

[0160] From the signal segments of the initial signal, select multiple signal segments whose inter-wave spacing values ​​reach the minimum inter-wave spacing threshold and do not exceed the maximum inter-wave spacing threshold.

[0161] In some optional embodiments, when the signal filtering module 43 filters multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value condition, it is specifically used for:

[0162] The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal.

[0163] Based on a preset scaling factor and the number of wave spacings in the initial signal, the number of signal segments to be deleted in the initial signal is determined as a first number.

[0164] In order of increasing inter-wave spacing, a second number of signal segments are deleted from at least one signal segment in the initial signal whose inter-wave spacing value is less than the average inter-wave spacing value.

[0165] According to the order of wave spacing from largest to smallest, a third number of signal segments are deleted from at least one signal segment in the initial signal whose wave spacing value is greater than the average spacing value; wherein, the sum of the second number and the third number is the first number;

[0166] The remaining signal segments are identified as the multiple signal segments obtained through filtering.

[0167] In some optional embodiments, before obtaining the heart rate of the target object based on the heartbeat fluctuation signal, the signal filtering module 43 is further configured to divide the heartbeat fluctuation signal into multiple first signals based on the position of the peak or trough in the heartbeat fluctuation signal, each first signal containing at least one cycle of signal.

[0168] The mean value of the first spacing corresponding to each segment of the first signal is calculated based on the spacing between each two adjacent peaks or troughs in each segment of the first signal.

[0169] The average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal.

[0170] Select the first signal whose first spacing mean is closest to the second spacing mean from each segment of the first signal as the reference signal;

[0171] For each segment of the first signal other than the reference signal, calculate the similarity between the first signal and the reference signal;

[0172] By splicing the reference signal and at least one segment of the first signal whose similarity to the reference signal meets a preset condition, a new heartbeat fluctuation signal is obtained.

[0173] In some alternative embodiments, each segment of the first signal has the same number of peaks and the same number of troughs.

[0174] In some optional embodiments, the images in the target image sequence contain regions of interest for a face, which include at least one of the following regions: a forehead region, a left cheek region, and a right cheek region;

[0175] The image acquisition module 41 is specifically used for:

[0176] Based on the video stream obtained from the target object during the first time period, a target image sequence consisting of multiple video frames containing the target object is obtained;

[0177] The signal extraction module 42 is specifically used for:

[0178] For each video frame in the target image sequence, the following processing is performed: based on face detection of the video frame, a face detection box of the target object is obtained; based on face region detection of the face region in the face detection box, face key points of the target object are determined; based on the face key points, the region of interest of the face contained in the video frame is determined; the sum of the average pixel values ​​of each region in the region of interest of the face is calculated as the signal value of the video frame.

[0179] A timing signal is obtained based on multiple sequentially consecutive signal values ​​within the first time period;

[0180] The blood volume pulse (rPPG) signal is extracted from the time-series signal and used as the initial signal.

[0181] In some optional embodiments, the heart rate determination module 44 is specifically used for:

[0182] The heart rate of the target object is selected from the multiple frequencies by performing a Fourier transform on the heart rate fluctuation signal.

[0183] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0184] This disclosure also provides an electronic device, such as... Figure 5 As shown, the electronic device includes a memory 51 and a processor 52. The memory 51 is used to store computer instructions that can be run on the processor, and the processor 52 is used to implement the heart rate detection method according to any embodiment of this disclosure when executing the computer instructions.

[0185] This disclosure also provides a computer program product, which includes a computer program / instructions that, when executed by a processor, implement the heart rate detection method described in any embodiment of this disclosure.

[0186] This disclosure also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the heart rate detection method described in any embodiment of this disclosure.

[0187] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of the solution in this specification according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0188] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0189] Other embodiments of this specification will readily occur to those skilled in the art upon consideration of the specification and practice of the invention claimed herein. This specification is intended to cover any variations, uses, or adaptations that follow the general principles of this specification and include common knowledge or customary techniques in the art not claimed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this specification are indicated by the following claims.

[0190] It should be understood that this specification is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this specification is limited only by the appended claims.

[0191] If the technical solution disclosed herein involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

[0192] The above description is merely a preferred embodiment of this specification and is not intended to limit this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

Claims

1. A heart rate detection method, characterized by, The method includes: Obtain a target image sequence, wherein the target image sequence contains multiple frames of images of the target object; Based on the target image sequence, an initial signal characterizing the heartbeat fluctuation of the target object is extracted; From the signal segments of the initial signal, multiple signal segments whose wave spacing values ​​meet the spacing value conditions are selected, and the multiple signal segments are spliced ​​together to obtain a heartbeat fluctuation signal composed of a stable signal. The signal segment is the signal between two adjacent wave peaks or the signal between two adjacent wave troughs in the initial signal; each signal segment represents a period of signal. The heart rate of the target object is obtained based on the heartbeat fluctuation signal; Before obtaining the heart rate of the target object based on the heartbeat fluctuation signal, the method further includes: Based on the position of the peak or trough in the heartbeat fluctuation signal, the heartbeat fluctuation signal is divided into multiple first signals, each of which contains at least one cycle of signal. The mean value of the first spacing corresponding to each segment of the first signal is calculated based on the spacing between each two adjacent peaks or troughs in each segment of the first signal. The average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal. Select the first signal whose first spacing mean is closest to the second spacing mean from each segment of the first signal as the reference signal; For each segment of the first signal other than the reference signal, calculate the similarity between the first signal and the reference signal; By splicing the reference signal and at least one segment of the first signal whose similarity to the reference signal meets a preset condition, a new heartbeat fluctuation signal is obtained.

2. The method according to claim 1, characterized in that, Before selecting multiple signal segments from the signal segments of the initial signal whose interval values ​​meet the interval value condition, and splicing the multiple signal segments to obtain the heartbeat fluctuation signal, the method further includes: Determine the signal between every two adjacent peaks and troughs in the initial signal, and the signal between every two adjacent troughs and peaks; The amplitudes of the signals between every two adjacent peaks and troughs, and between every two adjacent troughs and peaks, are scaled to the same range to obtain an initial signal with uniform amplitude.

3. The method according to claim 1, characterized in that, The step of filtering multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value conditions includes: The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal. Subtract a preset error value from the average spacing to determine the minimum threshold for wave spacing; add the preset error value to the average spacing to determine the maximum threshold for wave spacing. From the signal segments of the initial signal, select multiple signal segments whose inter-wave spacing values ​​reach the minimum inter-wave spacing threshold and do not exceed the maximum inter-wave spacing threshold.

4. The method according to claim 1, characterized in that, The step of filtering multiple signal segments from the signal segments of the initial signal whose spacing values ​​meet the spacing value conditions includes: The average spacing is calculated based on the spacing between every two adjacent peaks in the initial signal, or the average spacing is calculated based on the spacing between every two adjacent troughs in the initial signal. Based on a preset scaling factor and the number of wave spacings in the initial signal, the number of signal segments to be deleted in the initial signal is determined as a first number. In order of increasing inter-wave spacing, a second number of signal segments are deleted from at least one signal segment in the initial signal whose inter-wave spacing value is less than the average inter-wave spacing value. According to the order of wave spacing from largest to smallest, a third number of signal segments are deleted from at least one signal segment in the initial signal whose wave spacing value is greater than the average spacing value; wherein, the sum of the second number and the third number is the first number; The remaining signal segments are identified as the multiple signal segments obtained through filtering.

5. The method of claim 1, wherein, The number of peaks and the number of troughs are the same in each segment of the first signal.

6. The method according to claim 1, characterized in that, The images in the target image sequence contain regions of interest for a face, and the regions of interest for a face include at least one of the following regions: the forehead region, the left cheek region, and the right cheek region; The acquisition of the target image sequence includes: Based on the video stream obtained from the target object during the first time period, a target image sequence consisting of multiple video frames containing the target object is obtained; The step of extracting the initial signal characterizing the heartbeat fluctuation of the target object based on the target image sequence includes: For each video frame in the target image sequence, the following processing is performed: based on face detection of the video frame, a face detection box of the target object is obtained; based on face region detection of the face region in the face detection box, face key points of the target object are determined; based on the face key points, the region of interest of the face contained in the video frame is determined; the sum of the average pixel values ​​of each region in the region of interest of the face is calculated as the signal value of the video frame. A timing signal is obtained based on multiple sequentially consecutive signal values ​​within the first time period; The blood volume pulse (rPPG) signal is extracted from the time-series signal and used as the initial signal.

7. The method according to any one of claims 1 to 6, characterized in that, The step of obtaining the heart rate of the target object based on the heartbeat fluctuation signal includes: The heart rate of the target object is selected from the multiple frequencies by performing a Fourier transform on the heart rate fluctuation signal.

8. A heart rate detection device, characterized in that, The device includes: The image acquisition module is used to acquire a target image sequence, wherein the target image sequence contains multiple frames of images of the target object; The signal extraction module is used to extract an initial signal representing the heartbeat fluctuation of the target object based on the target image sequence; The signal filtering module is used to filter multiple signal segments from the signal segments of the initial signal whose wave spacing values ​​meet the spacing value conditions, and splice the multiple signal segments to obtain a heartbeat fluctuation signal composed of a stable signal. The signal segment is the signal between two adjacent wave peaks or the signal between two adjacent wave troughs in the initial signal; each signal segment represents a period of signal. A heart rate determination module is used to obtain the heart rate of the target object based on the heartbeat fluctuation signal; Before obtaining the heart rate of the target object based on the heartbeat fluctuation signal, the method further includes: Based on the position of the peak or trough in the heartbeat fluctuation signal, the heartbeat fluctuation signal is divided into multiple first signals, each of which contains at least one cycle of signal. The mean value of the first spacing corresponding to each segment of the first signal is calculated based on the spacing between each two adjacent peaks or troughs in each segment of the first signal. The average second spacing is calculated based on the spacing between every two adjacent peaks or troughs in the heartbeat fluctuation signal. Select the first signal whose first spacing mean is closest to the second spacing mean from each segment of the first signal as the reference signal; For each segment of the first signal other than the reference signal, calculate the similarity between the first signal and the reference signal; By splicing the reference signal and at least one segment of the first signal whose similarity to the reference signal meets a preset condition, a new heartbeat fluctuation signal is obtained.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store computer instructions that can be executed on the processor, and the processor being used to implement the heart rate detection method according to any one of claims 1 to 7 when executing the computer instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the heart rate detection method according to any one of claims 1 to 7.

Images