Line calibration and compensation method and system for multi-wavelength laser PPG sensor

Abstract

The present disclosure provides a line calibration and compensation method and system for a multi-wavelength laser PPG sensor. The multi-wavelength laser PPG sensor comprises a window and a plurality of lasers and a receiving channel inside the window. The line calibration method comprises: reading the equipment identification information of the multi-wavelength laser PPG sensor; measuring the dark state basic parameters of the receiving channel; controlling the output of the plurality of lasers to measure and establish the emission characteristic data of each laser; controlling the plurality of lasers to light up, collecting the signals reflected by the standard reflecting medium through the receiving channel, and calculating the channel normalization parameters; sequentially and individually lighting up one of the plurality of lasers, reading the residual response of other time slots or other paths in the receiving channel through the receiving channel to obtain the structure crosstalk data; and associating and packaging the dark state basic parameters, the emission characteristic data, the channel normalization parameters and the structure crosstalk data with the equipment identification information as an individualized parameter set.

Description

Technical Field

[0001] This disclosure relates to the field of testing, and in particular to a production line calibration and compensation method and system for a multi-wavelength laser PPG sensor. Background Technology

[0002] Existing photoplethysmography (PPG) wearable devices typically undergo LED brightness, photodetector (PD) dark current, analog front-end (AFE) gain, ambient light subtraction, and overall device functionality checks before leaving the factory.

[0003] For LED reflective PPGs, batch variations mainly stem from LED luminous flux, package angle, PD response, and assembly tolerances. Mass production variations for multi-wavelength laser PPGs are more complex: the center wavelength, threshold current, power-current profile, temperature drift, spot shape, and emission angle of 520nm, 660nm, and 940nm lasers may all differ; the same PD may also respond differently to different wavelengths; and assembly variations in light-blocking walls, windows, and skin-mounted structures can lead to different back-reflection and crosstalk baselines.

[0004] Existing multi-wavelength PPG or blood oxygen estimation solutions typically focus on runtime algorithms, light source placement, or wearable structures, rather than measuring multi-wavelength lasers, PDs, AFEs, window reflections, and structural crosstalk on a single production line and writing the results into the device.

[0005] If only batch average parameters are used, the same algorithm may produce different channel ratios and R values ​​(i.e., the ratio of the pulsation index of the red light channel to the pulsation index of the infrared light channel) on different devices.

[0006] The above information is provided as background information only and does not imply that the above information constitutes prior art in this disclosure. Summary of the Invention

[0007] One of the objectives of this disclosure is to provide a production line calibration method and system that can improve the consistency of data across wavelength channels of a multi-wavelength laser PPG sensor.

[0008] One of the objectives of this disclosure is to provide an operational compensation method that can eliminate inconsistencies in measurement results caused by differences in the manufacturing process of multi-wavelength laser PPG sensors.

[0009] According to a first aspect of this disclosure, a production line calibration method for a multi-wavelength laser PPG sensor is provided. The multi-wavelength laser PPG sensor includes a window and multiple lasers and a receiving channel located inside the window. The production line calibration method includes: reading the device identification information of the multi-wavelength laser PPG sensor; measuring the dark-state basic parameters of the receiving channel; controlling the output of multiple lasers to measure and establish emission characteristic data of each laser; controlling the multiple lasers to light up, acquiring signals reflected by a standard reflective medium through the receiving channel, and calculating channel normalization parameters; sequentially lighting up one of the multiple lasers individually, reading residual responses in other time slots or other paths in the receiving channel through the receiving channel to obtain structural crosstalk data; and associating the dark-state basic parameters, emission characteristic data, channel normalization parameters, and structural crosstalk data with the device identification information and encapsulating them into an individualized parameter set.

[0010] Optionally, one of the multiple lasers is lit up individually in sequence, and residual responses in other time slots or other paths in the receiving channel are read through the receiving channel to obtain structural crosstalk data. This may include: lighting up the first laser among the multiple lasers individually, and collecting background residual electrical signals in the non-emitting state through the receiving channel in the remaining time slots that are different from the lighting period of the first laser; and / or, when lighting up the first laser individually, collecting physical light leakage signals caused by optical path coupling through the remaining non-corresponding receiving channel paths; and generating structural crosstalk data corresponding to the first laser based on the background residual electrical signals and / or physical light leakage signals.

[0011] Optionally, the multiple lasers may include a second laser, a third laser, and a first laser. The first laser may be a green laser, the second laser may be a red laser, and the third laser may be an infrared laser. One of the multiple lasers is lit up individually in sequence, and the residual response in other time slots or other paths in the receiving channel is read through the receiving channel to obtain structural crosstalk data. The method also includes generating structural crosstalk data for the second and third lasers in the same way as generating structural crosstalk data for the first laser.

[0012] Optionally, controlling the output of multiple lasers to measure and establish emission characteristic data for each laser may include: sequentially illuminating multiple lasers and controlling each laser to output at multiple current points, pulse width points, or duty cycle points; measuring at least one of the center wavelength, main peak position, and spectral width of each laser, and measuring the output power of each laser at different outputs; and establishing emission characteristic data based on at least one of the center wavelength, main peak position, and spectral width of each laser, as well as the output power of each laser at different outputs.

[0013] Optionally, emission characteristic data can be established based on at least one of the center wavelength, main peak position, and spectral width of each laser, as well as the output power of each laser at different outputs. This may include: establishing a power drive lookup table for each laser based on the mapping relationship between output power and current point, pulse width point, or duty cycle point; establishing a safety upper limit for each laser, including maximum allowable drive current, maximum allowable pulse width, or maximum allowable duty cycle, based on the measured output power or center wavelength and a preset safety threshold; and establishing emission characteristic data based on the power drive lookup table and the safety upper limit.

[0014] Optionally, multiple lasers are controlled to light up, and signals reflected by a standard reflective medium are collected through a receiving channel to calculate channel normalization parameters. This may include: placing the standard reflective medium at a window; controlling multiple lasers to light up in the same time-division sequence during operation; collecting reflected signals through a receiving channel; and calculating channel normalization coefficients for each wavelength as channel normalization parameters. The channel normalization parameters are used to correct inconsistencies in the photoelectric conversion response of the receiving channel at different wavelengths.

[0015] Optionally, the channel normalization parameters may include: wavelength-dependent gain compensation values ​​for correcting channel sensitivity bias and offset compensation values ​​for correcting channel baseline offset.

[0016] Optionally, the multi-wavelength laser PPG sensor may also include non-volatile memory, and the production line calibration method further includes writing an individualized set of parameters into the non-volatile memory.

[0017] According to a second aspect of this disclosure, an operational compensation method for a multi-wavelength laser PPG sensor is provided. The operational compensation method includes: when the multi-wavelength laser PPG sensor is started, loading an individualized parameter set obtained according to the above-described production line calibration method; controlling the driving parameters of multiple lasers according to the individualized parameter set to output a target optical power; acquiring the original reflection signals of multiple receiving channels; and correcting the original reflection signals in real time according to the channel normalization parameters in the individualized parameter set to output a PPG signal.

[0018] According to a third aspect of this disclosure, a production line calibration system for a multi-wavelength laser PPG sensor is provided, comprising: a test environment module for fixing the multi-wavelength laser PPG sensor and constructing a standard reflected optical path; a signal measurement module for acquiring test data of the multi-wavelength laser PPG sensor; a processor connected to the signal measurement module; and a memory storing computer programs or instructions, wherein when the computer programs or instructions are executed by the processor, the production line calibration method described above is implemented.

[0019] The production line calibration method according to embodiments of this disclosure can achieve alignment of performance parameters of different individuals at the time of manufacture by measuring dark state basic data and establishing emission characteristic data, removing sensor noise and laser drive fluctuations.

[0020] The production line calibration method according to embodiments of the present disclosure can improve the signal-to-noise ratio and enhance the purity of multi-wavelength measurement data without altering the physical structure of the multi-wavelength laser PPG sensor.

[0021] The production line calibration method according to the embodiments of this disclosure associates and encapsulates core parameters such as dark-state basic data, emission characteristic data, channel normalization parameters, and structural crosstalk data with equipment identification information into an individualized parameter set, thereby realizing the mapping of sensor performance configuration and enabling the terminal to actively compensate for individual characteristics, ensuring the monitoring accuracy of the product. Attached Figure Description

[0022] The above and other objects and features of exemplary embodiments of this disclosure will become clearer from the following description taken in conjunction with the accompanying drawings, which exemplarily illustrate the embodiments, wherein: Figure 1 This is an overall flowchart of a production line calibration method according to an embodiment of the present disclosure.

[0023] Figure 2 This is a flowchart illustrating the specific process of acquiring structural crosstalk data according to embodiments of the present disclosure.

[0024] Figure 3 This is a flowchart illustrating the specific process of acquiring emission characteristic data according to an embodiment of the present disclosure.

[0025] Figure 4 This is a flowchart illustrating the specific process of obtaining channel normalization parameters according to embodiments of this disclosure; and Figure 5 This is a block diagram of a production line calibration system according to an embodiment of the present disclosure. Detailed Implementation

[0026] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings, examples of which are illustrated in the drawings, wherein the same reference numerals always refer to the same parts. The embodiments will now be described with reference to the accompanying drawings in order to explain this disclosure.

[0027] The following detailed description is provided to aid in obtaining a full understanding of the methods, apparatus, and / or systems described herein. However, the order of operations described herein is merely illustrative and is not limited to those orders set forth herein; equivalent substitutions or changes may be made, except for operations that must occur or be performed in a specific order. Furthermore, for clarity and conciseness, descriptions of content well-known in the art will be omitted or simplified.

[0028] Unless otherwise specified, the same reference numerals generally refer to the same elements (e.g., components, steps, and methods). Reference numerals described in previous embodiments that reappear in later embodiments may be omitted. Furthermore, technical features described in different or the same embodiments can be combined in any way, as long as the combined embodiment or technical solution is complete and can solve the technical problems of this application or achieve the technical effects described or not described in this disclosure but which can be determined based on the complete technical solution described above.

[0029] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains upon understanding this disclosure. Unless expressly defined herein, terms (such as those defined in a general dictionary) shall be interpreted as having a meaning consistent with their meaning in the context of the relevant field and in this disclosure, and shall not be interpreted in an idealized or overly formalistic manner. The terminology of this disclosure is described below.

[0030] Multi-wavelength laser PPG sensor: A precision optical detection module based on the principle of photoplethysmography (PPG). The multi-wavelength laser PPG sensor uses multiple lasers as emission sources. By measuring changes in the intensity of reflected or transmitted laser signals in human skin tissue, it achieves non-invasive monitoring of blood flow volume pulsation and physiological parameters. The multiple lasers in the multi-wavelength laser PPG sensor can emit lasers of different wavelengths. The production line calibration method disclosed herein is applicable to various multi-wavelength laser PPG sensors. The multi-wavelength laser PPG sensor may include a laser source array (e.g., 520nm, 660nm, and 940nm lasers) for emitting specific physiological monitoring beams, a sensitive photodetector (PD) for collecting tissue-reflected light, an analog front-end (AFE) responsible for signal amplification and conversion, and a window covering the outer layer of the module, etc. The receiving channel (or receiving path) of the multi-wavelength laser PPG sensor is a sensing link for capturing physiological reflected light and converting it into a digitally quantized signal, and may include a photodetector (PD), an analog front-end (AFE), and related signal transmission paths.

[0031] PD dark current: refers to the weak current generated by the thermal excitation of semiconductor materials in a photodetector (PD) under zero light (no light) conditions.

[0032] AFE bias: refers to the DC voltage deviation introduced by the AFE during signal amplification and digitization due to imperfect circuit components or power supply fluctuations, also known as "zero drift".

[0033] Noise level: refers to the sum of random noise generated by all electronic devices in the receiving channel in a silent state.

[0034] Baseline for each gain level: refers to the corresponding zero-level output of the AFE under different amplification settings.

[0035] Center wavelength: refers to the center of the energy distribution of the laser output spectrum, usually expressed as wavelength (unit: nanometer, nm).

[0036] Main peak position: refers to the wavelength corresponding to the point of maximum intensity in the energy distribution of the laser spectrum.

[0037] Spectral width: refers to the wavelength span corresponding to half the peak power of a laser's spectral energy distribution.

[0038] Power-driven lookup table: This is power-current, power-pulse width, or power-duty cycle relationship data, which records the correspondence between the laser's input electrical parameters and its output optical power.

[0039] Temperature drift: The unexpected shift in various core performance indicators of the sensor (including laser emission power, center wavelength, PD dark current, and AFE bias) caused by fluctuations in ambient temperature.

[0040] Reflection response at each wavelength: The original reflected signal intensity value collected by the PD in the receiving channel when each laser channel is lit sequentially.

[0041] Channel normalization coefficient: A weighted coefficient calculated by comparing the reflection response of each wavelength with a preset target range.

[0042] Transmitter-receiver crosstalk: refers to the non-target signal that is directly coupled to the receiver through the internal structure (e.g., encapsulation material, circuit board) after the laser is emitted from the transmitter without being reflected by the skin tissue.

[0043] Window back-reflection baseline: refers to the invalid light signal that is directly fed back to the detector when the laser shines on the outer glass window interface due to optical refraction / reflection.

[0044] The production line calibration method disclosed herein incorporates the measured wavelength, power-current relationship, threshold current, and temperature coefficient of the laser channel into the production line calibration.

[0045] The production line calibration method disclosed herein uses the equipment identification information of a multi-wavelength laser PPG sensor as a basis, measures the multi-dimensional calibration data of the multi-wavelength laser PPG sensor, and encapsulates the multi-dimensional calibration data together with the equipment identification information into an individualized parameter set, thereby realizing the correlation between the equipment identification information and the multi-dimensional calibration data.

[0046] The production line calibration method disclosed herein measures transmit-receive crosstalk, window back-baseline, etc. during the production line stage, which can obtain the noise floor reference of each physical channel. This enables the multi-wavelength laser PPG sensor to accurately extract the real physiological waveform from the mixed signal in various wearing environments, thereby improving the signal-to-noise ratio of physiological signals from the source.

[0047] The motion compensation method disclosed herein uses a personalized parameter set (e.g., a parameter table) for laser drive compensation, signal normalization, temperature compensation, crosstalk compensation, quality scoring, and safety threshold correction, thereby ensuring the safety of the laser output and correcting reference offsets in physiological signals. Furthermore, the motion compensation method of this disclosure allows the terminal algorithm to obtain an ideal measurement reference without complex development and adaptation for individual sensors. The accompanying drawings of this disclosure are described in detail below.

[0048] Figure 1 This is an overall flowchart of a production line calibration method according to an embodiment of the present disclosure. Figure 2 This is a flowchart illustrating the specific process of acquiring structural crosstalk data according to embodiments of the present disclosure. Figure 3 This is a flowchart illustrating the specific process of acquiring emission characteristic data according to embodiments of this disclosure. Figure 4 This is a flowchart illustrating the specific process of obtaining channel normalization parameters according to embodiments of the present disclosure.

[0049] The production line calibration method according to embodiments of this disclosure may include a step of reading device identification information, a calibration step, and a step of associating packaging.

[0050] The production line calibration method according to embodiments of the present disclosure may include steps S100, S200, S300, S400, S500 and S600.

[0051] In step S100, the device identification information of the multi-wavelength laser PPG sensor (hereinafter referred to as the sensor) is read.

[0052] Device identification information is used to uniquely identify the individual multi-wavelength laser PPG sensor, trace its physical properties at the time of manufacture, and associate it with its individualized calibration parameter set.

[0053] As an example, device identification information may include at least one of the following: unique serial number (SN), hardware version and batch information, laser batch, PD batch, and firmware version.

[0054] A unique serial number is a globally unique digital identifier assigned to a sensor.

[0055] Hardware version and batch information identify the iterative versions of the sensor hardware design and the batch sequence produced on the production line.

[0056] A laser batch refers to the production batch record of the laser emitting source integrated inside the sensor. Due to the discreteness of photoelectric properties in the manufacturing process of semiconductor lasers, there are slight differences in the center wavelength, luminous efficiency, and current-power response characteristics between different batches of light sources.

[0057] PD batch refers to the production source and batch information of photodetector (PD) components.

[0058] Firmware version refers to the version number of the control program running in the sensor's built-in processor. The firmware version defines the module's underlying timing logic, signal sampling strategy, and initial communication protocol.

[0059] In one example, during the initialization phase, specific read commands can be sent to the sensor register via various physical communication interfaces. The sensor then returns a data frame containing information such as a unique serial number, hardware batch, and firmware version. After verifying the integrity of the data, the main control unit in the production line calibration station indexes and loads the corresponding individualized calibration parameter set based on the identification information.

[0060] The production line calibration method disclosed herein can perform various calibrations, such as dark-state calibration, emission characteristic calibration, channel normalization calibration, and crosstalk calibration. As an example, the production line calibration method disclosed herein can also perform temperature drift calibration, etc.

[0061] In step S200, the dark-state fundamental parameters of the receiving channel are measured. The dark-state fundamental parameters may include at least one of PD dark current, AFE bias, noise level, and baseline for each gain level.

[0062] In one example, the AFE can be set to its highest sensitivity operating mode in a dark room where external light is completely blocked from entering the sensor, eliminating any excitation effect of ambient light on the photodetector. In this case, the weak current output by the PD is the PD dark current. As an example, the PD dark current can be obtained by integrating and quantizing the current using a precision galvanometer or the high-impedance conversion module within the AFE.

[0063] In one example, under conditions of complete darkness without any input signal, the simulated front-end circuitry at its static operating point can be used to read the DC voltage value at the output of the AFE, which is the AFE bias.

[0064] In one example, the receiving channel can be continuously sampled for an extended period while maintaining light shielding and a constant AFE configuration to acquire a set of time-series data. The noise level is then obtained by calculating the root mean square (RMS) or peak-to-peak value of the sequence.

[0065] In one example, because the AFE has a dynamic gain adjustment function, it is necessary to traverse all preset gain levels under the light-shielded condition. For each level, the signal acquisition and processing flow is repeated, and the corresponding digital output reference value is recorded.

[0066] In step S200, multiple lasers are controlled to output, so as to measure and establish emission characteristic data for each laser. The emission characteristic data may include at least one of the following: center wavelength, output power, drive current-power response curve, etc.

[0067] Reference Figure 2 The steps of controlling the output of multiple lasers to measure and establish emission characteristic data of each laser may include steps S310, S320 and S330.

[0068] In step S310, multiple lasers are lit sequentially. Specifically, multiple lasers can be lit sequentially, and each laser can be controlled to output at multiple current points, pulse width points, or duty cycle points.

[0069] In step S320, at least one of the center wavelength, main peak position and spectral width of each laser is measured, and the output power of each laser is measured under different output conditions.

[0070] As an example, in a fully shaded environment, to obtain accurate, individualized emission characteristics of lasers, sensors can be securely mounted using automated fixtures, ensuring precise alignment of each laser's window with the optical axis of the light signal acquisition system. Subsequently, a connection is established via a communication interface between the main control unit (e.g., the main control unit of a computer device in a production line calibration station) to ensure the sensors are in a programmable state. During measurement, each laser can be activated sequentially according to a pre-defined test sequence.

[0071] For a single laser, the master control unit sends precise control signals, which are then used by the drive module to traverse a predefined space of combinations of current, pulse width, and duty cycle. After each set of parameters is switched, a stabilization wait time can be introduced to ensure that the laser enters a steady state.

[0072] During this period, the spectrometer acquires and analyzes optical signals in real time through an optical fiber probe, accurately recording the center wavelength, main peak position, and full width at half maximum (FWHM) of the laser under the current driving conditions. Specifically, multiple lasers (e.g., 520nm, 660nm, and 940nm lasers) can be driven sequentially, and the center wavelength, main peak position, or spectral width can be measured using the spectrometer module, recording the measured values. Simultaneously, the optical power meter installed at the end of the optical path uses an integrating sphere to homogenize the diverging beam, eliminating spatial inhomogeneity interference, thereby measuring the output power under this driving combination.

[0073] In step S330, emission characteristic data is established. Specifically, emission characteristic data can be established based on at least one of the center wavelength, main peak position, and spectral width of each laser, as well as the output power of each laser at different outputs.

[0074] For example, a power drive lookup table for each laser can be established based on the mapping relationship between output power and current point, pulse width point, or duty cycle point; a safety upper limit for each laser, including maximum allowable drive current, maximum allowable pulse width, or maximum allowable duty cycle, can be established based on the measured output power or center wavelength and preset safety thresholds; and emission characteristic data can be established based on the power drive lookup table and the safety upper limit.

[0075] In one example, the collected test data can be converted into a response matrix of current-pulse width-duty cycle-output power. During processing, the power-drive current mapping sequence can be extracted for each configuration using the duty cycle and pulse width as indices, thus constructing multiple LI (current-power) curves under different pulse configurations.

[0076] In one example, the threshold current of the laser can be calculated, and then the slope efficiency can be calculated using least squares or piecewise linear fitting. For intervals with strong nonlinearity, interpolation algorithms (such as linear interpolation or spline interpolation) can be used to fill the response data between sampling points, thereby obtaining a wide-range lookup table.

[0077] In one example, a lookup table containing power output values, corresponding current compensation amounts, and pulse duty cycle adjustment coefficients can be constructed using current as the primary index.

[0078] Specifically, the output power of each laser can be measured at multiple current points, pulse width points, or duty cycle points to establish power-current, power-pulse width, or power-duty cycle relationships, thereby obtaining a power drive lookup table and a safety upper limit.

[0079] In one example, the functional relationships between current, pulse width, duty cycle, and output light energy can be derived first using previously measured driving conditions and power response models. Then, based on the maximum permissible radiation thresholds set for different wavelengths in relevant laser safety standards, the critical physical parameters under the current individual laser characteristics can be obtained through reverse calculation. For each laser, the corresponding maximum permissible driving current, maximum pulse width, and duty cycle, etc., can be calculated.

[0080] The power drive lookup table records the mapping relationship between drive parameters (current, pulse width, duty cycle) and optical output (power, wavelength), enabling the system to precisely adjust the laser's output level according to monitoring requirements. The safety upper limit, based on laser safety standards and device tolerance, specifies the limit values ​​for the drive parameters. The power drive lookup table and the safety upper limit together constitute emission characteristic data.

[0081] In step S400, multiple lasers are controlled to light up, and signals reflected by the standard reflective medium are collected through the receiving channel to calculate the channel normalization parameters.

[0082] A standard reflective medium, tissue simulator, or optical phantom can be placed at the skin-contact window, and three laser channels can be lit according to the time-division sequence during operation. The reflection response of each wavelength can be measured to obtain the channel normalized parameters.

[0083] Specifically, a standard reflective medium can be placed at the window; multiple lasers can be controlled to light up in the same time-division sequence during operation; the reflected signals can be collected through the receiving channel, and the normalization coefficient of each wavelength channel can be calculated as the channel normalization parameter. This channel normalization parameter can be used to correct the inconsistency of photoelectric conversion response of the receiving channel at different wavelengths.

[0084] The channel normalization parameters here may include a gain correction factor. As an example, a channel with moderate performance can be selected as a reference (or the arithmetic mean of all channels can be used as the baseline response value). Then, for each channel, its scaling factor relative to the reference is calculated, i.e., the gain correction factor ( ).

[0085] Furthermore, the channel normalization parameters may also include wavelength-dependent offset compensation values ​​used to correct channel baseline offset. Specifically, the ideal reference zero point can be compared with the measured dark-state output value of a channel, and the channel-specific offset compensation amount can be obtained by calculating the difference between the two. However, this is merely an example, and the channel normalization parameters disclosed herein are not limited to this.

[0086] In step S500, structural crosstalk data calibration is performed. This calibration method transforms passive tolerance of crosstalk into active interference compensation. By measuring the residual response, a compensation model is constructed, thereby enabling the theoretical detection limit to be approached through algorithm-level correction without increasing hardware costs.

[0087] Specifically, one of multiple lasers can be lit up individually in sequence, and the residual response in other time slots or other paths in the receiving channel can be read through the receiving channel to obtain structural crosstalk data.

[0088] By reading the residual responses from other paths, a closed-loop model can be constructed, enabling the detector to accurately determine which signals are reflections from the real environment and which are internal interferences, regardless of where the light comes from, thus ensuring the reliability of the monitoring results.

[0089] By employing a single-point transmission and multi-point reception mode, the crosstalk between each transmission channel (Tx) and a specific reception channel (Rx) can be quantified, thereby establishing a crosstalk coupling matrix. The aforementioned crosstalk calibration can eliminate these unwanted optical signals, thus improving measurement sensitivity and anti-interference capability. It can automatically compensate for minor assembly errors generated during manufacturing (such as slight deviations in lens position). Even with minor differences in the physical structure of each device, the crosstalk calibration disclosed herein can acquire its unique crosstalk data, ensuring high consistency among different devices at the time of manufacture.

[0090] Reference Figure 3 The steps of sequentially lighting up one of multiple lasers individually and reading the residual responses in other time slots or other paths in the receiving channel to obtain structural crosstalk data may include steps S510, S520 and S530.

[0091] In step S510, a residual background electrical signal is acquired. For example, the first laser among multiple lasers can be lit individually, and the residual background electrical signal in the non-emitting state can be acquired through the receiving channel during the remaining time slots that are different from the lighting period of the first laser.

[0092] Taking the production line calibration station mentioned above as an example, the station includes a computer as the control core, which coordinates the scheduling of various components through a communication bus. During operation, a movable fixture precisely places the PPG module under test into the light-shielding box, ensuring the optical path is in a completely enclosed, dark environment. First, the computer in the calibration station issues a command to extinguish all lasers. During the non-emitting time slot, it directly reads and records the residual background electrical signal from the module's PD receiving channel to establish an internal crosstalk and noise baseline. Subsequently, the computer restarts the lasers, acquiring wavelength and power data through a spectrometer and calibration power meter for optical calibration. Additionally, a temperature sensor continuously monitors the module's thermal state and provides real-time feedback, offering necessary parameters for the final comprehensive performance deviation correction.

[0093] Additionally, the computer issues calibration commands, which drive the module to execute a specific time-division multiplexing (TDM) sequence. In the immediate time slot following the first laser's shutdown (or in other time slots where the channel is not illuminated), the computer keeps the probe link active. Within the logical time slot where the laser is off, the workstation computer instructs the module's AFE (analog front-end) to continuously acquire sampled values ​​via the PD. Since the laser is off, there should be no photon incidence; however, due to optical leakage within the structure (e.g., cover plate reflection, glue scattering), the PD will still output a weak residual electrical signal.

[0094] In step S520, physical light leakage signals are collected. When the first laser is lit alone, physical light leakage signals generated by optical path coupling are collected through the other non-corresponding receiving channel paths.

[0095] As an example, the computer in the production line calibration station instructs the first laser (e.g., 520nm) to emit power at the rated drive current. Simultaneously, the station's computer instructs the AFE (Analog Front End) of the module to activate a non-corresponding receiving channel (i.e., a PD receiving channel other than the normal physical path corresponding to this laser). At this time, the PD receiving channel does not collect the target reflected signal, but rather the photoelectric conversion current generated after laser photons enter the PD through paths such as total internal reflection from the cover plate, internal scattering from the encapsulation adhesive, or light leakage within the structure. The computer can synchronously read the raw ADC count value of this specific receiving channel. Since this channel should be receiving signals from other time slots at this time, any non-zero value read can be mapped to the physical light leakage intensity of that path.

[0096] In step S530, structural crosstalk data is generated. This structural crosstalk data, corresponding to the first laser, can be generated based on the background residual electrical signal and / or physical light leakage signal. The background residual electrical signal and / or physical light leakage signal can be signals that have undergone normalization or averaging filtering to remove random noise.

[0097] As an example, the workstation computer can directly use the acquired background residual electrical signal as the noise floor reference, and combine it with the physical light leakage signal calculated according to a certain weight to generate a set of compensation offsets. When the module is actually working, these fixed data can be directly subtracted from the total signal acquired by the sensor.

[0098] As an example, the workstation computer can calculate the crosstalk ratio per unit current by subtracting the background residual reference from the physical light leakage signal and then dividing by the driving current used for measurement. The resulting structural crosstalk data is a mapping table containing the interference ratios between each channel. During runtime, the current interference value can be calculated and subtracted in real time based on the real-time laser driving current. This method can handle multi-power operating scenarios.

[0099] As an example, the workstation computer can collect background residual and physical light leakage signals under different ambient temperatures, and fit these data points to generate a compensation function that varies with temperature. The generated structural crosstalk data can be a mathematical model that includes temperature-sensitive parameters. The built-in temperature sensor values ​​can be read in real time during operation, and the model can be used to dynamically calculate the compensation amount adapted to the current thermal state.

[0100] As an example, the workstation computer repeats the above steps, switching between different combinations of transmitter and receiver channels in turn, traversing all potential coupling paths, and binding the collected values ​​with the current channel gain configuration and module ambient temperature, which can be recorded in the workstation computer's database.

[0101] Specifically, in addition to the first laser, the multiple lasers may also include a second laser and a third laser. The first laser can be a green laser, the second laser can be a red laser, and the third laser can be an infrared laser. The structural crosstalk data of the second and third lasers can be generated in the same way as the structural crosstalk data of the first laser.

[0102] However, this is just one example; different methods can be used to determine structural crosstalk data for each laser. In addition, crosstalk data can be determined through electronic circuit interference compensation, differential cancellation measurements (which utilize the consistency of inter-channel interference to directly weaken interference at its source by constructing ratios or differential logic), and machine learning intelligent fitting (which trains weights through massive amounts of data to handle complex, nonlinear structural light leakage).

[0103] During the production line phase, slight fluctuations in the reflectivity of the cover plate, the refractive index of the encapsulating adhesive, and the light transmittance of PPG modules are unavoidable between batches. Unlike methods that rely on post-production algorithms to filter out crosstalk, the production line calibration disclosed herein converts the interference into a simple lookup table (LUT) or coefficient matrix and burns it onto the device or other locations. During actual operation, the module only requires simple subtraction or multiplication calculations to eliminate the interference. This not only reduces the computational burden on the main control chip of terminal devices (e.g., smartwatches, wristbands) but also effectively reduces power consumption and extends battery life.

[0104] In addition to the above, structural crosstalk can also be calibrated in the following ways. For example, by applying specific modulation frequencies to different lasers, the corresponding interference frequency components of each channel can be directly extracted in the receiving channel using phase-locked loop or digital filtering techniques, thereby establishing a crosstalk model based on frequency characteristics. Alternatively, instead of directly subtracting the absolute value, the responses of the target path and the reference path can be simultaneously acquired, and the nonlinear change ratio between the two can be calculated. The common-mode characteristics of the interference under the same environment can then be used to cancel it out.

[0105] For the measurement and compensation of structural crosstalk, a combination of multi-dimensional physical experimental environments and different decision-making logics can be adopted. For example, in the physical measurement stage, a dark box environment is constructed to ensure the purity of the noise level acquisition; light-absorbing shielding components are used to block excess diffuse reflection and accurately extract leakage characteristics inside the optical path; standard reflective components are used to simulate the optical reflection characteristics of skin to ensure that the measurement benchmark conforms to real working conditions; and testing in a non-contact state can effectively reduce structural deformation interference caused by clamps and other pressure.

[0106] The subsequent compensation logic is as follows: interference coefficients from each path are removed from the original signal in real time to ensure the dynamic accuracy of the measurement data. Specifically, structural crosstalk data can be constructed into an interference transfer matrix for dynamic removal of the original signal during PPG signal acquisition. Simultaneously, a threshold determination mechanism is introduced to automatically intercept modules with crosstalk exceeding the expected range, preventing defective assembled products from entering the market at the source. Furthermore, the stability of the terminal monitoring output can be ensured by dynamically adjusting the weights of subsequent algorithms. As an example, the timing during the acquisition process can be consistent with the working timing of the PPG sensor in heart rate / blood oxygen monitoring mode.

[0107] Additionally, although not shown, temperature drift calibration can also be performed. Specifically, during the R&D phase, small batches of modules can be used for scanning to extract the patterns of physical light leakage and crosstalk matrix changes with temperature, and a general compensation model can be constructed. During the production phase, it can be simplified to single-unit room temperature calibration, collecting and calculating the individual reference matrix of the module, and then logically fusing the matrix with the batch temperature model to generate exclusive dynamic compensation parameters (which can then be burned into the module).

[0108] The calibration process of this disclosure has been described in detail above. The calibration process described below is described in conjunction with the calibration station on the production line.

[0109] First, the workstation computer can read the device identification information of the multi-wavelength laser PPG sensor in order to establish a unique identity association for the parameters generated subsequently.

[0110] Subsequently, the workstation computer controls the device under test (DUT) to enter a closed testing environment. With all lasers turned off, the background residual electrical signal is acquired through the receiving channel, i.e., the dark-state fundamental parameters of the receiving channel are measured to obtain the reference level of the sensor in a completely dark environment, including the dark current and bias values ​​at various gain levels of the analog front-end (AFE). By measuring the dark-state fundamental parameters, the static performance of the sensor circuit can be effectively evaluated. If the acquired dark-state noise value exceeds the preset safety threshold, the workstation computer will determine that the sensor has a potential dark current anomaly or circuit fault, and mark the device as requiring retesting, thereby ensuring that all subsequent optical signal calibrations are based on a clean electrical noise floor.

[0111] To obtain the precise output characteristics of each light source in the sensor, the workstation computer can sequentially illuminate multiple lasers and control each laser to output at multiple preset current points, pulse width points, or duty cycle points. During emission, a spectrometer and a calibrated optical power meter are used to measure at least one of the center wavelength, peak position, and spectral width of each laser, while simultaneously measuring its optical power under different output states. Based on these measured data, the workstation computer constructs a mapping relationship between output power and current, pulse width, or duty cycle points, thereby generating a power drive lookup table for each laser. Simultaneously, based on the measured center wavelength and output power, combined with preset safety thresholds, the safety upper limit for each laser is calculated and determined, i.e., the maximum allowable drive current, maximum allowable pulse width, or maximum allowable duty cycle. Finally, the above information is integrated and established into complete emission characteristic data, ensuring that the module meets both performance indicators and laser safety standards in subsequent applications.

[0112] During channel normalization calibration, the workstation computer controls a movable fixture to precisely place a standard reflective medium at the sensor window to simulate the real optical reflection environment of human skin. Simultaneously, the workstation computer controls multiple lasers to illuminate according to a time-division sequence identical to the actual operating environment, and the receiving channel collects the signals reflected by the standard reflective medium. Based on the collected feedback information, channel normalization parameters are calculated. These parameters include wavelength-related gain compensation values ​​to correct channel sensitivity deviations, and offset compensation values ​​to correct channel baseline offsets. This eliminates inconsistencies in the photoelectric conversion response of the receiving channel at different wavelengths.

[0113] The calculated channel normalization parameters ensure that the reflection response of different wavelengths falls within a uniform target range when the sensor faces a standard reflective medium, thereby eliminating measurement deviations caused by batch differences in devices or different spectral responses.

[0114] Next, when the workstation computer performs crosstalk testing, it sequentially and individually illuminates one of multiple lasers (e.g., a green laser, a red laser, or an infrared laser). While the first laser is illuminated, the workstation computer acquires signals through its receiving channels. Specifically, this includes acquiring background residual electrical signals in non-emitting states during other time slots that differ from the laser's illumination period, and / or acquiring physical light leakage signals caused by optical path coupling during the illumination period through other non-corresponding receiving channel paths. Based on the acquired background residual electrical signals and physical light leakage signals, the workstation computer generates structural crosstalk data corresponding to that laser, and then iterates through the remaining lasers in the same manner to generate structural crosstalk data.

[0115] In step S600, associated encapsulation is performed. Dark-state basic parameters, emission characteristic data, channel normalization parameters, and structural crosstalk data can be associated and encapsulated with device identification information into an individualized parameter set.

[0116] Specifically, the workstation computer can simultaneously collect the module's dark-state basic parameters (noise floor), emission characteristic data (light source driving characteristics), and normalized parameters of each channel, combining them with previously generated structural crosstalk data to form multi-dimensional indicators. Subsequently, by reading the unique identifier (UID) embedded in the module chip or associating it with production batch information, all the above parameters can be encoded according to a specific address mapping table to form an individualized parameter set in binary format. This individualized parameter set can be verified and burned into the module's non-volatile memory area (e.g., EEPROM), ensuring that the main control program can automatically retrieve the matching parameter set through the UID, achieving cross-device optical consistency compensation. However, this is just an example; it can also be written to other devices or locations, such as the sensor MCU, main control MCU, external EEPROM, secure storage area, or cloud database, but the device side only needs to retain the local parameters required for operation.

[0117] Reference Figure 4 The operation compensation method disclosed herein may include steps S1000, S2000, S3000 and S4000.

[0118] In step S1000, when the multi-wavelength laser PPG sensor is started, the above-mentioned individualized parameter set is loaded.

[0119] When the multi-wavelength laser PPG sensor enters motion monitoring or user-activated state, it can immediately trigger the invocation of an individualized parameter set. By reading the device identification information (UID), the system can match a uniquely corresponding dark-state basic parameter, emission characteristic data, channel normalization parameter, and structural crosstalk data.

[0120] In step S2000, the driving parameters of multiple lasers are controlled according to an individualized parameter set to output the target optical power.

[0121] After obtaining the individualized parameter set, closed-loop control can be implemented on the drive circuits of multiple lasers based on the power drive lookup table and safety upper limit recorded therein.

[0122] For example, for lasers of different wavelengths such as red, green, and infrared, the optimal drive current, pulse width, and duty cycle values ​​can be automatically mapped from a table based on the target optical power requirements. Furthermore, the drive load can be monitored in real time to ensure that all outputs are strictly limited within the threshold of the "maximum permissible drive parameters" defined by the individualized parameter set. This guarantees both the linearity and stability of the light intensity output at each wavelength while strictly adhering to laser safety regulations.

[0123] In step S3000, the raw reflected signals from multiple receiving channels are acquired.

[0124] While the laser emits light according to a preset time-division multiplexing (TDM) sequence, the receiving channel simultaneously initiates sampling. This not only captures the effective reflected signal at the target wavelength but also performs signal isolation based on individualized parameter sets. Since the reflected signal is superimposed with ambient light interference and physical coupling within the optical path, multi-channel parallel sampling technology can be used to record the level data in real time across multiple receiving paths. The resulting raw reflected signal constitutes the core noise floor and dynamic components of the PPG waveform.

[0125] In step S4000, the original reflected signal is corrected in real time according to the channel normalization parameters in the individualized parameter set to output the PPG signal.

[0126] Specifically, the channel normalization parameters and structural crosstalk baseline from the individualized parameter set can be called to perform real-time correction on the acquired raw reflection signal.

[0127] For example, dark-state fundamental parameters can be used to eliminate circuit static noise, and the sensitivity deviation of different wavelengths can be uniformly calibrated by channel normalization parameters (gain and bias compensation values).

[0128] Subsequently, the system incorporates structural crosstalk data to precisely remove physical light leakage and residual background electrical signals caused by optical path coupling from the original signal, achieving deep decoupling of crosstalk noise. Through this compensation, the original signal, which was contaminated by assembly tolerances and optical interference, is reconstructed into a pure PPG signal. This eliminates individual device differences and also improves the signal-to-noise ratio of the PPG waveform through real-time calculation.

[0129] While various calibration methods have been described above, this disclosure is not limited to the examples mentioned above. For instance, a multi-order crosstalk compensation model can be established for nonlinear optical feedback. Considering that structural crosstalk may not change linearly with the driving current, this disclosure can also introduce a quadratic term correction model or a lookup table (LUT) interpolation algorithm, that is, based on the measured values ​​at at least two different driving current points, a compensation function containing nonlinear coefficients is fitted to achieve accurate correction of crosstalk under different luminous intensities.

[0130] Secondly, to eliminate the impact of charge injection and signal tailing during time slot switching of the AFE analog front end on the measurement reference, the acquisition of background residual electrical signals and physical light leakage signals can be limited to a specific window period when the signals tend to be in a steady state. By avoiding the initial transient range after the ADC is turned on within this time slot, it is ensured that the acquired residual electrical signals fully reflect the current component generated by optical feedback, thus avoiding misjudging the circuit response as structural crosstalk.

[0131] Finally, an ambient background light suppression scheme can be introduced, that is, adding an ambient background reflection measurement before illuminating the laser. For example, with the laser off, the photocurrent baseline value under the current tooling environment can be collected and subtracted from the subsequent physical leakage signal measurement value. This effectively decouples the internal optical feedback (physical leakage) of the module from the external environmental reflection, ensuring that the generated structural crosstalk data only characterizes the individual characteristics of the module's internal assembly errors.

[0132] The production line calibration system 5000 disclosed herein can be the aforementioned production line calibration station. The test environment module 5100, as the physical basis, can integrate a light shield, a movable fixture, a standard diffuse reflector, and a tissue simulator, and can be used to fix the equipment and construct a standard reflected light path.

[0133] Reference Figure 5 The production line calibration system 5000 disclosed herein may include a test environment module 5100, a signal measurement module 5200, a processor 5300, and a memory 5400.

[0134] The Test Environment Module 5100 is used to fix the multi-wavelength laser PPG sensor and construct a standard reflected optical path. The Test Environment Module 5100 can fix the device under test through a precision mechanical fixture, and according to calibration requirements, a light shield, a standard diffuse reflector, or a tissue simulator can be moved to the sensor window position to ensure a high degree of consistency of the optical test environment and stability of the physical optical path.

[0135] The signal measurement module 5200 can be used to acquire test data of the multi-wavelength laser PPG sensor. Under the scheduling of the processor 5300, this module uses a spectrometer, a calibration optical power meter, and a temperature sensor to acquire in real time the spectral data, optical power data, temperature characteristic data, and the original electrical reflection response under a standard reflective medium of the sensor under different laser driving currents and pulse widths.

[0136] The processor 5300 can be connected to the signal measurement module. The processor 5300 can call program instructions in the memory 5400 to execute complex algorithms such as reading device identification information, calculating dark state basic parameters, establishing transmission characteristic data, calculating channel normalization parameters, and extracting structural crosstalk data, and can format and encapsulate the final results.

[0137] The Memory 5400 can store preset thresholds, safety drive limit parameters, multi-temperature compensation models, and the final generated individual parameter set required for the calibration process, and provides highly reliable data read and write support for the processor, ensuring that calibration parameters can be safely written into the non-volatile memory of the device under test.

[0138] The signal measurement module 5200 may include a spectrometer, a calibration optical power meter, and a temperature sensor, and is responsible for acquiring the module's emission, reflection, and temperature data in real time.

[0139] The processor 5300 and memory 5400 together form the core of the workstation computer. The former acts as the scheduling center to execute calibration algorithms and logical operations, while the latter stores calibration programs, preset thresholds, and the final individualized parameter set.

[0140] Each module works collaboratively through the scheduling of the processor 5300, integrating the originally scattered optical performance testing and parameter configuration into an automated production process. This ensures that each PPG module can complete calibration from structural crosstalk decoupling and channel normalization to temperature drift compensation, and finally leaves the factory with its own personalized parameter set.

[0141] The production line calibration method according to embodiments of this disclosure can improve the data consistency of each wavelength channel of a multi-wavelength laser PPG sensor.

[0142] The production line calibration method according to embodiments of the present disclosure can eliminate inconsistencies in measurement results caused by differences in the manufacturing processes of multi-wavelength laser PPG sensors.

[0143] The production line calibration method according to embodiments of this disclosure can achieve alignment of performance parameters of different individuals at the time of manufacture by measuring dark state basic data and establishing emission characteristic data, removing sensor noise and laser drive fluctuations.

[0144] The production line calibration method according to embodiments of the present disclosure can improve the signal-to-noise ratio and enhance the purity of multi-wavelength measurement data without altering the physical structure of the multi-wavelength laser PPG sensor.

[0145] The production line calibration method according to the embodiments of this disclosure associates and encapsulates core parameters such as dark-state basic data, emission characteristic data, channel normalization parameters, and structural crosstalk data with equipment identification information into an individualized parameter set, thereby realizing the mapping of sensor performance configuration and enabling the terminal to actively compensate for individual characteristics, ensuring the monitoring accuracy of the product.

[0146] While specific terminology has been used to describe various embodiments of this disclosure, the specification and drawings are to be regarded as illustrative rather than restrictive in order to aid in understanding this disclosure. Various modifications and changes can be made by those skilled in the art (e.g., different features in different embodiments may be combined) without departing from the broader spirit and scope of this disclosure. Therefore, the scope of this disclosure is not limited by the specific embodiments and examples, but by the claims and their equivalents.

Claims

1. A production line calibration method for a multi-wavelength laser PPG sensor, wherein the multi-wavelength laser PPG sensor includes a window and multiple lasers and receiving channels located inside the window, characterized in that, include: Read the device identification information of the multi-wavelength laser PPG sensor; Measure the basic dark-state parameters of the receiving channel; Control the output of the multiple lasers to measure and establish emission characteristic data for each laser; The system controls the lighting of the multiple lasers, collects signals reflected by a standard reflective medium through the receiving channel, and calculates the channel normalization parameters. One of the plurality of lasers is lit up individually in sequence, and the residual response in other time slots or other paths in the receiving channel is read through the receiving channel to obtain structural crosstalk data; The dark state basic parameters, the emission characteristic data, the channel normalization parameters, and the structural crosstalk data are associated with and encapsulated into an individualized parameter set with the device identification information.

2. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 1, characterized in that, By sequentially and individually illuminating one of the plurality of lasers, and reading the residual responses in other time slots or other paths within the receiving channel through the receiving channel, structural crosstalk data is obtained, including: The first laser among the plurality of lasers is individually illuminated, and during the remaining time slots that differ from the illumination period of the first laser, the background residual electrical signal in the non-emitting state is acquired through the receiving channel; and / or, When the first laser is lit alone, the physical light leakage signal caused by optical path coupling is collected through the other non-corresponding receiving channel paths; Based on the background residual electrical signal and / or physical light leakage signal, structural crosstalk data corresponding to the first laser is generated.

3. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 2, characterized in that, The plurality of lasers includes a second laser, a third laser, and a first laser, wherein the first laser is a green laser, the second laser is a red laser, and the third laser is an infrared laser. Sequentially and individually illuminating one of the plurality of lasers, and reading residual responses in other time slots or other paths of the receiving channel through the receiving channel to obtain structural crosstalk data, the method further includes: The structural crosstalk data for the second and third lasers are generated in the same manner as the structural crosstalk data for the first laser.

4. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 1, characterized in that, Controlling the output of the multiple lasers to measure and establish emission characteristic data for each laser, including: The plurality of lasers are lit sequentially, and each laser is controlled to output at multiple current points, pulse width points, or duty cycle points; Measure at least one of the center wavelength, main peak position and spectral width of each laser, and measure the output power of each laser at different outputs; Based on at least one of the center wavelength, main peak position, and spectral width of each laser, as well as the output power of each laser at different outputs, emission characteristic data are established.

5. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 4, characterized in that, Based on at least one of the center wavelength, main peak position, and spectral width of each laser, as well as the output power of each laser at different outputs, emission characteristic data are established, including: Based on the mapping relationship between the output power and the current point, pulse width point or duty cycle point, a power drive lookup table is established for each laser. Based on the measured output power or center wavelength and the preset safety threshold, establish a safety upper limit for each laser, including the maximum allowable drive current, the maximum allowable pulse width, or the maximum allowable duty cycle. Based on the power-driven lookup table and the safety upper limit, launch characteristic data is established.

6. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 1, characterized in that, Controlling the illumination of the multiple lasers, acquiring signals reflected by the standard reflective medium through the receiving channel, and calculating channel normalization parameters, including: Place a standard reflective medium at the window; The multiple lasers are controlled to illuminate according to the same time-division sequence during operation. Reflected signals are acquired through the receiving channel, and the normalization coefficient for each wavelength channel is calculated as the channel normalization parameter. The channel normalization parameter is used to correct the inconsistency in the photoelectric conversion response of the receiving channel at different wavelengths.

7. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 6, characterized in that, The channel normalization parameters include: wavelength-dependent gain compensation values ​​used to correct channel sensitivity deviations and offset compensation values ​​used to correct channel baseline offsets.

8. The production line calibration method for the multi-wavelength laser PPG sensor according to claim 1, characterized in that, The multi-wavelength laser PPG sensor also includes a non-volatile memory, and the production line calibration method further includes: writing the individualized parameter set into the non-volatile memory.

9. A method for operational compensation of a multi-wavelength laser PPG sensor, characterized in that, include: When the multi-wavelength laser PPG sensor is started, an individualized parameter set obtained by the production line calibration method according to any one of claims 1 to 8 is loaded; The driving parameters of the plurality of lasers are controlled according to the individualized parameter set to output the target optical power; Collect raw reflected signals from multiple receiving channels; The original reflected signal is corrected in real time based on the channel normalization parameters in the individualized parameter set to output a PPG signal.

10. A production line calibration system for multi-wavelength laser PPG sensors, characterized in that, include: The test environment module is used to fix the multi-wavelength laser PPG sensor and construct a standard reflected optical path; The signal measurement module is used to acquire test data from the multi-wavelength laser PPG sensor. The processor is connected to the signal measurement module; A memory storing a computer program or instructions that, when executed by the processor, implement the production line calibration method according to any one of claims 1 to 8.

Images