Blood oxygen detection method based on wide-band synchronous excitation light source and multispectral imaging

Abstract

The present invention belongs to the technical field of spectral component detection and relates to a blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral images. The method comprises the following steps: Step S1: a spectral band separation algorithm; based on the spectral characteristics of blood oxygen, skin, and ambient light, a narrower operating band is selected within the entire light source spectral range, and multispectral images are extracted for the blood oxygen-sensitive band, the skin-sensitive band, and the skin-sensitive band; Step S2: a skin-quality and skin-color interference removal algorithm; based on the band images extracted by the spectral band separation algorithm in the wavelength range of 400nm to 700nm, skin-quality and skin-color are identified, and errors in blood oxygen detection caused by differences in skin-quality and skin-color are corrected; Step S3: a synchronous PPG signal extraction and blood oxygen calculation algorithm; based on the spectral band separation algorithm, images of different characteristic bands are obtained from the same multispectral image, corresponding PPG signals are extracted from the different band images, and blood oxygen values ​​are calculated. The advantages of the present invention are: high system integration, compact size, and strong portability.

Description

Technical Field

[0001] The present invention relates to a non-invasive blood oxygen detection method, and more particularly to a blood oxygen detection method based on a wide-band synchronous excitation light source and multi-spectral imaging. Background Art

[0002] With the continuous development of society and economy, people are paying more and more attention to their physical condition. Among them, blood oxygen saturation is an important parameter indicator, which is mainly used to reflect the body's oxygen supply capacity. Currently, there are some methods for non-invasively monitoring the real-time blood oxygen status of the human body based on optical detection. For example, in the Chinese patent publication number CN 113598761 A, a dual-wavelength infrared blood oxygen detection system based on CCD is disclosed. This invention uses a near-infrared point light source to time-share two wavelengths of light sources, and the CCD sensor collects image information, extracts the PPG signal and calculates the blood oxygen value. For example, in the Chinese patent publication number CN 11355861A, a dual-wavelength photoacoustic blood oxygen detection method is disclosed. Two pulsed laser light sources of different wavelengths are alternately triggered to obtain ultrasonic waves, and the ultrasonic waves are analyzed to obtain the blood oxygen value. For example, in the Chinese patent publication number CN 114073520 A, a green light-based blood oxygen detection device, a blood oxygen detection method and a medium thereof are disclosed. The detection device includes a PPG sensor, and the PPG sensor includes at least one green light source, a red light source, at least one infrared light source and at least one photoelectric sensor. This invention judges the quality of the PPG signal by adding a green light signal, thereby improving the accuracy of blood oxygen detection. For example, in the Chinese patent CN 114847944 Document A describes a hyperspectral cerebral blood oxygen detection system. The system includes a controller, light source, light source detector, filter turntable, data acquisition, and data processing modules. Multiple filters enable the extraction and application of light in different wavelength bands. The paper "Research on In Vivo Tissue Detection Methods Based on Multispectral Imaging" uses characteristic narrowband filters as spectroscopic elements to build a multispectral imaging system, combining Wiener estimation and the second-order derivative of the spectrum to calculate blood oxygen levels. For example, Chinese patent CN 115089171 describes a method for enhancing the accuracy of blood oxygen detection results in people with dark skin. This patent utilizes an oximeter MCU to receive the RGB values ​​of skin color detected by a color sensor. The detected RGB values ​​determine whether to use the first light-emitting diode (normal power) or the second light-emitting diode (increased power) for blood oxygen detection. The paper "Research on Non-Contact Blood Oxygen Saturation Detection Technology Based on IPPG" proposes a non-contact blood oxygen calculation method. This method primarily involves recording a video of a person's face, extracting physiological information from the image, separating the channels to obtain R and B channel data, and then filtering and transforming the data to calculate the blood oxygen level.

[0003] Based on the above patents, they can be divided into two categories: one is the time-sharing detection type, and the other is the multispectral and hyperspectral type. The time-sharing detection type includes patents: CN 113598761 A A CCD-based dual-wavelength infrared blood oxygen detection system, CN11355861 A dual-wavelength photoacoustic blood oxygen detection method, etc., which have the advantages of small size and low cost, but time-sharing detection is easily interfered by electromagnetic waves in the environment, and the detection accuracy is low; multispectral and hyperspectral types, including patent CN 114847944A A hyperspectral cerebral blood oxygen detection system, the paper "Research on in vivo tissue detection method based on multispectral imaging", etc., have the advantages of high detection accuracy and good real-time performance, but have problems such as large size, complex structure, high cost, and poor adaptability to different skin types and skin colors.

[0004] There is an urgent need for a blood oxygen detection system based on a wide-band synchronous excitation light source and multispectral imaging to solve the problems of low accuracy of time-sharing detection; large size of multispectral and hyperspectral blood oxygen detection equipment, complex detection system, high cost and poor portability. Summary of the Invention

[0005] In response to the above-mentioned problems, the present invention provides a blood oxygen detection system and method based on a wide-band synchronous excitation light source and multispectral imaging. These methods achieve synchronous PPG signal extraction and blood oxygen calculation, resolving issues such as signal interference during time-sharing acquisition. Furthermore, the wide-band synchronous excitation light source offers advantages such as wide spectral coverage, high information content, and strong applicability. This method enables the extraction of band images corresponding to different bands from the same multispectral image, as well as the synchronous extraction of PPG signals for blood oxygen value calculation. Furthermore, it corrects blood oxygen detection errors caused by differences in skin quality and complexion, improving blood oxygen detection accuracy. These methods overcome the shortcomings of the aforementioned prior art.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral imaging, characterized by comprising the following steps:

[0008] Step S1: spectrum segment separation algorithm;

[0009] According to the spectral characteristics of blood oxygen, skin, and ambient light, a narrower working band is selected within the entire light source spectrum, and multispectral images of the same bands of blood oxygen sensitive bands, skin sensitive bands, and skin color sensitive bands are extracted;

[0010] Step S2: remove skin texture and skin color interference algorithm;

[0011] The wavelength band images in the range of 400nm to 700nm extracted based on the spectrum separation algorithm are used to identify skin texture and skin color, and correct the error in blood oxygen detection caused by differences in skin texture and skin color;

[0012] Step S3: Synchronously extract the PPG signal and calculate the blood oxygen algorithm;

[0013] Based on the spectral band separation algorithm, images of different characteristic bands are obtained from the same multispectral image, and the corresponding PPG signals are extracted from the images of different bands to calculate the blood oxygen value.

[0014] As a preferred embodiment of the present invention, the spectrum segment separation algorithm includes the following steps:

[0015] Step S11: pre-store spectral information of commonly used blood oxygen and skin characteristic bands as reference spectral information, recorded as light_spec;

[0016] Step S12: Turn on the broadband synchronous excitation light source, use the multispectral imaging chip to collect a multispectral image of the fingertip, perform background and noise removal preprocessing on the collected multispectral image, and use the preprocessed multispectral image as the original image, recorded as original_img;

[0017] Step S13: Based on the fact that each pixel in original_img obtained in step S12 is a superposition of pixels of images in each band, the pixel expression of the multispectral image is as follows;

[0018] x(i,j)=∑A×λ(i,j) (1)

[0019] Where: x(i,j) is the pixel data of original_img obtained in step S12; λ(i,j) is the pixel information of each band image; A is the proportion of the pixel information of each band image in the pixel data of original_img, that is, ∑A=A1+A2+A3+...=1;

[0020] Step S14: Calculate the distance between the original_img pixel data and the fitted pixel data according to formula (2). When the distance is less than 10 -4 When , output A;

[0021] dis = ||x(i,j)-A×λ(i,j)|| (2)

[0022] Step S15: Based on step S13, the proportion of pixel information of different bands in the multispectral image is solved. Combined with the spectral information light_spec of different bands in step S11, the spectrum corresponding to each band can be obtained according to formula (3);

[0023] spinv =light_spec × A (3)

[0024] Step S16: Based on the spectral information obtained in step S15, a desired band is selected and summed to obtain a multispectral image of the corresponding band;

[0025] img = ∑spinv(λ). (4)

[0026] As a preferred embodiment of the present invention, the algorithm for removing skin quality and skin color interference includes the following steps:

[0027] Step S21: extracting the band images corresponding to the wavelength range of 400nm to 700nm according to the spectrum separation algorithm;

[0028] Step S22: Obtain the corresponding spectrum curve according to formula (3);

[0029] Step S23: pre-store and set the spectrum curve corresponding to yellow skin as the skin quality and skin color standard curve, recorded as spec; calculate the deviation between the spectrum curve obtained in step S22 and the standard curve; calculate the deviation, recorded as bias;

[0030] bias = 1 / n×∑(spec - spinv) 2 (5)

[0031] Where: n is the number of spectral range intervals.

[0032] As a preferred embodiment of the present invention, the algorithm for synchronously extracting PPG signals and calculating blood oxygen comprises the following steps:

[0033] Step S31: Obtaining visible light and near-infrared 940nm images based on a spectral separation algorithm, extracting corresponding PPG signals based on the visible light and near-infrared 940nm images, and simultaneously extracting direct current (DC) and alternating current (AC) signals and calculating the amplitudes of the two signals;

[0034] Step S32: Obtain the AC signal amplitude (I AC ) and DC signal amplitude (I DC ) ratio;

[0035] A light =I AC / I DC (6)

[0036] Among them: I AC is the amplitude of the AC signal in the PPG signal; I DC is the amplitude of the DC signal in the PPG signal; light is the ratio of the AC signal amplitude to the DC signal amplitude in the PPG signal corresponding to the same band;

[0037] Step S33: Calculate the light corresponding to each band according to formula (6), and calculate the R value according to formula (7);

[0038] R = Alight1 / Alight2 (7)

[0039] Step S34: Correct the skin quality and skin color deviation according to formula (8):

[0040] R xz = R × bias (8)

[0041] Where: R xz It is the R value after correcting the influence of skin texture and skin color, and bias is the deviation between the spectrum of the skin texture and skin color to be tested and the spectrum of the standard skin texture and skin color;

[0042] Step S35: calibrate and correct the blood oxygen value affected by skin color differences according to formula (9):

[0043] SpO2 = C × R xz 2 + D × R xz + E (9)

[0044] Where: SpO2 is the blood oxygen saturation value, C, D and E are calibration coefficients, R xz It is the R value after correcting the influence of skin texture and skin color.

[0045] As a preferred embodiment of the present invention, the wide-spectrum synchronous excitation light source is used to emit light of a continuous spectrum covering the blood oxygen and skin sensitive spectrum, and the spectral range is 350nm-950nm.

[0046] As a preferred embodiment of the present invention, the multispectral imaging chip is used to collect multispectral images of the fingertips.

[0047] Another object of the present invention is to provide a blood oxygen detection system based on a wide-band synchronous excitation light source and multispectral imaging, comprising a wide-band synchronous excitation light source, a multispectral imaging chip, and an algorithm processing module. The wide-band synchronous excitation light source is configured to emit light of a continuous spectrum. The multispectral imaging chip collects multispectral images of the fingertips. The algorithm processing module comprises a spectral separation algorithm module, a skin quality and skin color interference removal algorithm module, and a synchronous PPG signal extraction and blood oxygen calculation algorithm module. The spectral separation algorithm module selects a narrower operating band within the entire light source spectrum based on the spectral characteristics of blood oxygen, skin, and ambient light, and extracts multispectral images of different bands sensitive to blood oxygen, skin quality, and skin color, thereby eliminating various interferences and calculating accurate blood oxygen values. The skin quality and skin color interference removal algorithm module identifies skin quality and skin color based on the visible light band images extracted by the spectral separation algorithm module, and corrects the error in blood oxygen detection caused by skin quality and skin color deviation. The synchronous PPG signal extraction and blood oxygen calculation algorithm module extracts PPG signals corresponding to different bands at the same time on the same multispectral image and calculates the blood oxygen value.

[0048] As a preferred embodiment of the present invention, the wide-spectrum synchronous excitation light source is used to emit light of a continuous spectrum covering the blood oxygen and skin sensitive spectrum, and the spectral range is 350nm-950nm.

[0049] The advantages and positive effects of the present invention are: the problem of time-sharing triggering of blood oxygen detection by two light sources of different wavelengths in the prior art is solved by using a wide-band synchronous excitation light source, which shortens the blood oxygen detection time, reduces the detection error, and improves the measurement accuracy; at the same time, the wide-band synchronous excitation light source is combined with a spectrum separation algorithm to realize the selection of a narrower working band within the entire light source spectrum according to the spectral characteristics of blood oxygen, skin, ambient light, etc., and to realize the extraction of multispectral images of different bands such as blood oxygen sensitive bands, skin quality sensitive bands, and skin color sensitive bands, so as to realize the extraction of visible light band images for skin quality and skin Color recognition corrects the error in blood oxygen detection caused by skin quality and skin color differences, extracts visible light band and near-infrared band images for blood oxygen detection, reduces the complexity of the detection system, and reduces the size of the detection device; based on the synchronous extraction of PPG signals and calculation of blood oxygen algorithm, it avoids the error caused by different blood states at different times and improves the accuracy of blood oxygen detection; therefore, the present invention is based on the wide-band synchronous excitation light source to generate multiple spectrum bands and a large amount of information, synchronously extracts PPG signals and calculates blood oxygen, thereby improving the accuracy of blood oxygen detection; at the same time, the system of the present invention has high integration, small size, and strong portability. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] By referring to the following description in conjunction with the accompanying drawings, and with a more complete understanding of the present invention, other objects and results of the present invention will become more clear and easy to understand. In the accompanying drawings:

[0051] Figure 1 This is a schematic diagram of the blood oxygen detection system structure based on a wide-band synchronous excitation light source and multi-spectral imaging;

[0052] Figure 2 It is the flow chart of the spectrum segment separation algorithm;

[0053] Figure 3 It is the flow chart of the algorithm for removing skin interference;

[0054] Figure 4 This is a flow chart of a blood oxygen detection system based on a wide-band synchronous excitation light source and multi-spectral imaging;

[0055] Description of the drawings: finger to be detected 1, wide-band synchronous excitation light source 2, integrated module 3, circuit interface 4, multispectral imaging chip 5, red blood cells 6. Implementation Method

[0056] In the following description, for illustrative purposes, numerous specific details are set forth to provide a comprehensive understanding of one or more embodiments. However, it will be apparent that the embodiments may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form to facilitate description of one or more embodiments. Example 1

[0057] Figure 1 FIG. 2 shows a schematic diagram of the overall structure according to an embodiment of the present invention.

[0058] A blood oxygen detection system based on a wide-band synchronous excitation light source and multispectral imaging, comprising a hardware part and an algorithm processing module. The hardware part mainly includes: a wide-band synchronous excitation light source 1 and a multispectral imaging chip 5; the algorithm processing module is integrated in an integrated module 3, and a circuit interface 4 is installed on one side of the integrated module 3, wherein the algorithm processing module includes: a spectrum separation algorithm module, a skin quality and skin color interference removal algorithm module, and a synchronous PPG signal extraction and blood oxygen calculation algorithm module; the wide-band synchronous excitation light source 2 can emit light in a continuous wavelength band covering blood oxygen and skin-sensitive spectrum bands; the multispectral imaging chip 5 is mainly used to perform multispectral image acquisition on the fingertip area irradiated by the wide-band synchronous excitation light source 2. Its working spectrum band can cover blood oxygen and skin-sensitive spectrum bands, and the spectral range is 350nm-950nm, recording exposure time and time series; the multispectral imaging chip 2 used in this embodiment is patent number 202010211115.0, and the patent name is a multispectral imaging chip and color imaging method capable of realizing color imaging. Example 2

[0059] like Figure 1-4 As shown, this example specifically includes the following steps:

[0060] Step S1: For the red blood cells 6 in the finger 1 to be detected, the fingertip of the finger 1 to be detected is pressed on the integrated module 3 equipped with a broadband synchronous excitation light source 2 and a multispectral imaging chip 5;

[0061] Step S2: The broadband synchronous excitation light source 2 is turned on to vertically illuminate the fingertip of the finger to be detected 1;

[0062] Step S3: The multispectral imaging chip 5 starts collecting multispectral images of the fingertip of the finger to be detected 1, and saves the exposure time and time series.

[0063] Step S4: the multispectral imaging chip 5 ends the acquisition and simultaneously controls the wide-band synchronous excitation light source 2 to be turned off;

[0064] Step S5: The spectrum separation algorithm saves the spectrum information of different bands used in advance, which is recorded as: light_spec;

[0065] Step S6: preprocessing the multispectral image collected by the multispectral imaging chip 5, and using the preprocessed image as raw data;

[0066] Step S7: The spectrum separation algorithm calculates the proportion of pixels corresponding to the visible light band of the multispectral image collected by the multispectral imaging chip 5 according to formula (1), and obtains the proportion of the visible light band image pixels in the original image pixels;

[0067] x(i,j)=∑A×λ(i,j) (1)

[0068] Where: x(i,j) is the pixel data of the original image obtained by S6; λ(i,j) is the pixel information of each band image; A is the proportion of the pixel information of each band image in the original image, that is, ∑A=A1+A2+A3+...=1;

[0069] Step S8: Set the initial value of A, and use the iterative calculation method to calculate the distance between the original image pixel and the fitted pixel. When the distance is less than 10 -4 When , the output corresponds to a proportion A;

[0070] dis = ||x(i,j)-A×λ(i,j)|| (2)

[0071] Step S9: Based on the corresponding optimal pixel ratio obtained in step S8, we can obtain the corresponding visible spectrum information;

[0072] spinv =light_spec × A (3)

[0073] Step S10: Compare the visible light spectrum obtained in step S9 with the set standard spectrum, denoted as spec, and calculate the bias value according to formula (5):

[0074] bias = 1 / n×∑(spec - spinv) 2 (5)

[0075] Step S11: The spectrum separation algorithm calculates the spectra of the visible light band and the near infrared 940nm wavelength according to formula (1), formula (2) and formula (3);

[0076] Step S12: Based on the spectrum information of the visible light band and the near infrared 940nm wavelength calculated in step S11, the corresponding multispectral image is calculated according to formula (4);

[0077] img = ∑spinv(λ) (4)

[0078] Step S13: Extract the PPG signal based on the visible light image and the image corresponding to the near-infrared wavelength of 940nm obtained in step S12; calculate Alight1 corresponding to the visible light band and Alight2 corresponding to the near-infrared wavelength of 940nm according to formula (6).

[0079] A light =I AC / I DC (6)

[0080] Step S14: Calculate the R coefficient according to the Alight calculated in step S13 according to formula (7);

[0081] R = Alight1 / Alight2 (7)

[0082] Step S15: Based on the skin quality and skin color deviation bias calculated in step S10, it is necessary to correct the skin quality and skin color deviation. Then, the corrected R is obtained according to formula (8): xz ;

[0083] R xz = R × bias (8)

[0084] Step S16: Calculate the blood oxygen value under the skin type and skin color according to formula (9):

[0085] SpO2 = C × R xz 2 + D × R xz + E (9)

[0086] Where: SpO2 is the blood oxygen saturation value, C, D and E are calibration coefficients, Rxz It is the R value after correcting the influence of skin texture and skin color.

[0087] In summary, the present invention, based on a wide-band synchronous excitation light source and multispectral imaging, realizes that the wide-band synchronous excitation light source method of measuring blood oxygen replaces the existing dual-wavelength dual-light source alternating flashing measurement method; it broadens the band range, enriches the spectral information, reduces measurement errors, shortens image acquisition time, avoids errors such as baseline drift and potential electric interference, and improves the accuracy of blood oxygen detection.

[0088] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A blood oxygen detection method based on a broadband synchronous excitation light source and multispectral imaging, characterized by: The following steps are involved: Step S1: spectrum segment separation algorithm; According to the spectral characteristics of blood oxygen, skin, and ambient light, a narrower working band is selected within the entire light source spectrum, and multispectral images of the same bands of blood oxygen sensitive bands, skin sensitive bands, and skin color sensitive bands are extracted; Step S2: remove skin texture and skin color interference algorithm; The wavelength band images in the range of 400nm to 700nm extracted based on the spectrum separation algorithm are used to identify skin texture and skin color, and correct the error in blood oxygen detection caused by skin texture and skin color differences; Step S3: Synchronously extract the PPG signal and calculate the blood oxygen algorithm; Based on the spectrum separation algorithm, images of different characteristic bands are obtained from the same multispectral image, and the corresponding PPG signals are extracted from the images of different bands to calculate the blood oxygen value; The spectrum segment separation algorithm comprises the following steps: Step S11: pre-store spectral information of commonly used blood oxygen and skin characteristic bands as reference spectral information, recorded as light_spec; Step S12: Turn on the broadband synchronous excitation light source, use the multispectral imaging chip to collect a multispectral image of the fingertip, perform background and noise removal preprocessing on the collected multispectral image, and use the preprocessed multispectral image as the original image, recorded as original_img; Step S13: Based on the fact that each pixel in original_img obtained in step S12 is a superposition of pixels of images in each band, the pixel expression of the multispectral image is as follows; x(i,j)=∑A×λ(i,j) (1) Where: x(i, j) is the pixel data of original_img obtained in step S12; λ(i, j) is the pixel information of each band image; A is the proportion of the pixel information of each band image in the pixel data of original_img, that is, ∑A=A1+A2+A3+...=1; Step S14: Calculate the distance between the original_img pixel data and the fitted pixel data according to formula (2). -4 When , output A; dis = ||x(i,j)-A×λ(i,j)|| (2) Step S15: Based on step S13, the proportion of pixel information of different bands in the multispectral image is solved. Combined with the spectral information light_spec of different bands in step S11, the spectrum corresponding to each band can be obtained according to formula (3); spinv = light_spec × A (3) Step S16: Based on the spectral information obtained in step S15, a desired band is selected and summed to obtain a multispectral image of the corresponding band; img = ∑spinv(λ) (4).

2. The blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral imaging according to claim 1, characterized in that: The algorithm for removing skin quality and skin color interference includes the following steps: Step S21: extracting the band images corresponding to the wavelength range of 400nm to 700nm according to the spectrum separation algorithm; Step S22: Obtain the corresponding spectrum curve according to formula (3); Step S23: pre-store and set the spectrum curve corresponding to yellow skin as the skin quality and skin color standard curve, recorded as spec; calculate the deviation between the spectrum curve obtained in step S22 and the standard curve; Calculate the deviation, denoted as bias; bias = 1 / n×∑(spec - spinv) 2 (5) Where: n is the number of spectral range intervals.

3. The blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral imaging according to claim 1, characterized in that: The algorithm for synchronously extracting PPG signals and calculating blood oxygen comprises the following steps: Step S31: Obtaining visible light band and near-infrared 940nm images based on a spectral band separation algorithm, extracting corresponding PPG signals based on the visible light band and near-infrared 940nm images, and simultaneously extracting DC and AC signals and calculating the amplitudes of the two signals; Step S32: Obtaining the ratio of the AC signal amplitude to the DC signal amplitude of the PPG signal in different bands according to formula (6); A light =I AC / I DC (6) Where: I AC is the amplitude of the AC signal in the PPG signal; I DC is the amplitude of the DC signal in the PPG signal; A light It is the ratio of the AC signal amplitude to the DC signal amplitude in the PPG signal corresponding to the same band; Step S33: Calculate the A corresponding to each band according to formula (6) light , calculate the R value according to formula (7); R = A light1 / A light2 (7) Step S34: Correct the skin quality and skin color deviation according to formula (8): R xz = R × bias (8) Where: R xz It is the R value after correcting the influence of skin texture and skin color, and bias is the deviation between the spectrum of the skin texture and skin color to be tested and the spectrum of the standard skin texture and skin color; Step S35: calibrate and correct the blood oxygen value affected by skin color differences according to formula (9): SpO2 = C ×R xz 2 + D × R xz + E (9) Where: SpO2 is the blood oxygen saturation value, C, D and E are calibration coefficients, R xz It is the R value after correcting the influence of skin texture and skin color.

4. The blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral imaging according to claim 1, characterized in that: The wide-spectrum synchronous excitation light source is used to emit light of a continuous spectrum covering the blood oxygen and skin sensitive spectrum, and the spectral range is 350nm-950nm.

5. The blood oxygen detection method based on a wide-band synchronous excitation light source and multispectral imaging according to claim 1, characterized in that: The multispectral imaging chip is used to collect multispectral images of the fingertips.

6. A blood oxygen detection system based on a broadband synchronous excitation light source and multispectral imaging used in the method of claim 1, characterized in that: It includes a wide-spectrum synchronous excitation light source, a multi-spectral imaging chip, and an algorithm processing module; the wide-spectrum synchronous excitation light source is used to emit light of a continuous spectrum; The multispectral imaging chip collects multispectral images of the fingertips; The algorithm processing module includes: a spectrum separation algorithm module, a skin texture interference removal algorithm module, and a synchronous PPG signal extraction and blood oxygen calculation algorithm module; The spectral band separation algorithm module selects a narrower working band within the entire light source spectrum range according to the spectral characteristics of blood oxygen, skin, and ambient light, and extracts multispectral images of different bands of blood oxygen sensitive bands, skin quality sensitive bands, and skin color sensitive bands; the skin quality and skin color interference removal algorithm module identifies skin quality and skin color based on the visible light band image extracted by the spectral band separation algorithm module, and corrects the error of skin quality and skin color deviation caused by blood oxygen detection; the synchronous extraction of PPG signals and calculation of blood oxygen algorithm module is used to extract PPG signals corresponding to different bands at the same time on the same multispectral image and calculate the blood oxygen value.

7. A blood oxygen detection system based on a broadband synchronous excitation light source and multispectral imaging used in the method according to claim 6, characterized in that: The wide-spectrum synchronous excitation light source is used to emit light of a continuous spectrum covering the blood oxygen and skin sensitive spectrum, and the spectral range is 350nm-950nm.

Images