CN103263271A - Non-contact automatic blood oxygen saturation degree measurement system and measurement method - Google Patents

Abstract

The invention discloses a non-contact automatic blood oxygen saturation degree measurement system and measurement method. Measurement of a blood oxygen saturation degree is carried out through video shooting. The measurement system comprises a video collection module, a primary color component separation module, an ROI extraction module, a time-varying signal generation module, a filtering and noise reduction module and a signal calibration and analysis module. The measurement method comprises the steps: collecting a color video recording information of the wrist portion of a human body, extracting a red component image of the color video, extracting a region which includes the artery as ROI images, extracting a mean value of all pixels of each frame of ROI image as a feature value of the frame of ROI image, generating a time-amplitude waveform, conducting filtering and noise reduction on a waveform signal, conducting normalization and calibration on the filtered signal, then finding wave crests and wave troughs of the signal in a waveform period, obtaining a specific value of a wave crest value and a wave trough value, and estimating the blood oxygen saturation degree (SpO2) through the specific value. A non-contact remote physiological signal detection method is adopted, motion artifact impact in the video recording is overcome, measurement precision is high, and multiple people can be measured simultaneously.

Description

Contactless automatic oxygen saturation measurement system and measuring method

Technical field

The present invention relates to the electronic health care station equipment, particularly relate to a kind of system and method that carries out blood pressure measurement automatically.

Background technology

Traditional measurement method of blood oxygen saturation is divided into manual measurement and measures dual mode automatically.Wherein: manual measurement needs advanced pedestrian's body blood sampling, and the recycling blood gas analyzer carries out electro chemical analysis, according to the blood oxygen pressure PO that measures ₂Calculate blood oxygen saturation, this method is cumbersome, and can not carry out continuous monitoring.Automatically metering system is realized by multiple measurement device, for example can adopt the finger cot type photoelectric sensor: during measurement, need sensor sleeve is guaranteed that at finger the sensor tested human body of speaking keeps in touch, utilize finger as the transparent vessel of splendid attire hemoglobin, use the HONGGUANG of wavelength 660nm and the near infrared light conduct of 940nm to inject light source, measure the light conduction intensity by tissue bed, calculate hemoglobin concentration and blood oxygen saturation; If wear the discomfort that checkout gear may cause people's emotion changes and health for a long time.The remote measurement of blood oxygen saturation can provide more comfortable and relative accurate electrodeless physiology assessment.Up to the present, right and wrong automatization is nearly all attempted in the research and development that all contactless blood oxygen saturations are automatically measured, and in use is subject to the motion artifacts influence.

Summary of the invention

Problem based on above-mentioned prior art existence, the present invention proposes a kind of contactless automatic oxygen saturation measurement system and measuring method, this system is by the color video of record human body wrist information, the means that adopt primary colours to separate are got the red component image and are analyzed, can obtain metrical information after the analyzing and processing, even under the situation that motion artifacts exists, it still has degree of precision.

A kind of contactless automatic oxygen saturation measurement system that the present invention proposes, adopt video capture to carry out oxygen saturation measurement, this system comprises video acquisition module, primary color component separation module, ROI extraction module, time varying signal generation module, filtering and noise reduction module and signal calibration analysis module; Wherein:

Described video acquisition module is for the color video of the segment record human body wrist information of collection;

Described primary color component separation module is used for every two field picture of color video is carried out the primary colours separation, obtains wherein red component image;

Described ROI extraction module utilizes the texture information algorithm of wrist blood vessel and venation that above-mentioned every frame red component image is positioned, extract wherein comprise tremulous pulse in interior zone the ROI as this two field picture;

Described time varying signal generation module is used for getting the average of every all pixels of frame ROI image as the eigenvalue of this two field picture, generates tremulous pulse pulse amplitude time dependent the time change waveshape signal;

Described filtering and noise reduction module is used for above-mentioned waveshape signal being carried out filtering and noise reduction, filtering arterial pulse wave wherein and the interference of respiratory wave;

Described signal calibration analysis module is used for filtered waveshape signal " x (n) " is carried out normalized, and its formula is:

$\hat{x}\left(n\right)=\frac{x\left(n\right)-x_{\min }\left(n\right)}{x_{\max }\left(n\right)-x_{\min }\left(n\right)},$

X wherein _Min(n) be the minima of signal x (n), x _Max(n) be the maximum of signal x (n); And it is calibrated, find out crest value and the trough value of signal in a wave period then, obtain the ratio of trough value and crest value, utilize this ratio to estimate blood oxygen saturation SpO2; Computing formula is:

SpO2=kR/(1+kR)；

Wherein R is the ratio of trough value and crest value, and k is for adjusting operator.

If the color type of described color video is RGB, then directly video image is carried out primary colours and separate; If video image color type right and wrong RGB, then be translated into and carry out primary colours again behind the RGB color type and separate.

The invention allows for a kind of contactless automatic measurement method of blood oxygen saturation, this method may further comprise the steps:

Step 1 is gathered the color video that comprises a segment record human body wrist information;

Step 2 is carried out primary colours to the colored every two field picture of gathering that frequently carries out and is separated, and extracts the red component image;

Step 3 is utilized by the texture location algorithm red component image that obtains is carried out the ROI location, extract comprise tremulous pulse in interior zone the ROI image as this two field picture, comprise the texture information of wrist blood vessel and venation in the described red component image;

Step 4 is got the average of every all pixels of frame ROI image as the eigenvalue of this two field picture, generates the waveshape signal of a time dependent time-amplitude of tremulous pulse pulse amplitude;

Step 5 is carried out filtering and noise reduction to waveshape signal, the interference of filtering arterial pulse wave and respiratory wave;

Step 6 is carried out normalization to filtered waveshape signal x (n), and its formula is

$\hat{x}\left(n\right)=\frac{x\left(n\right)-x_{\min }\left(n\right)}{x_{\max }\left(n\right)-x_{\min }\left(n\right)},$

X wherein _Min(n) be the minima of signal x (n), x _Max(n) be the maximum of signal x (n); And it is calibrated, find out crest and the trough of signal in a wave period then, obtain the ratio of trough value and crest value, utilize this ratio to estimate blood oxygen saturation (SpO2)

S _pO ₂=kR/(1+kR)；

Wherein R is the ratio of trough value and crest value, and k is one and adjusts operator.

If the color type of described color video is RGB, then directly video image is carried out primary colours and separate; If video image color type right and wrong RGB, then be translated into and carry out primary colours again behind the RGB color type and separate.

Compared with prior art, the present invention is based on the human body wrist and comprise the color video of the texture information of wrist blood vessel and venation, utilization is from motion tracking and primary colours isolation technics, adopt the method for noinvasive, contactless remote detection physiological signal, overcome the influence of motion artifacts in the videograph, motion has been had tolerance, the certainty of measurement height, can carry out many people automatically simultaneously and measure, that obtain is the oxygen saturation measurement result.

Description of drawings

Fig. 1 is the system construction drawing of contactless automatic oxygen saturation measurement of the present invention system;

Fig. 2 is the flow chart of contactless automatic measurement method of blood oxygen saturation of the present invention.

The specific embodiment

Below in conjunction with accompanying drawing and preferred embodiment, to according to the specific embodiment provided by the invention, structure, feature and effect thereof, be described in detail as follows.

As shown in Figure 1, be the system construction drawing of contactless automatic oxygen saturation measurement of the present invention system.This system comprises video acquisition module 101, primary color component separation module 102, ROI extraction module 103, time varying signal generation module 104, filtering and noise reduction module 105 and signal calibration analysis module 106.

As shown in Figure 2, contactless automatic measurement method of blood oxygen saturation of the present invention comprises following flow process:

Employing has the common photographic head of automatic gain function or the color video that the higher special-purpose photographic head collection of resolution comprises people's wrist, step 201; The video of gathering is carried out branch frame primary colours separate, extract the red component image and (if the color type of video image is RGB, then directly image is carried out RGB and separate; If video image color type right and wrong RGB, then be translated into and carry out primary colours again behind the RGB color type and separate), step 202; Step 202; Utilization is carried out the ROI location by the texture location algorithm to the red component image that obtains (having comprised the wrist blood vessel that collects and the texture information of venation in the red component image that separates), extract comprise tremulous pulse in interior zone the ROI as this two field picture, step 203;

Every frame ROI image is got the average of all pixels respectively, as the eigenvalue of this two field picture, generate the time signal wave that becomes of corresponding time-amplitude, step 204; The signal that obtains is carried out filtering and noise reduction, i.e. the interference of filtering arterial pulse wave and respiratory wave, step 205; After above-mentioned processing finishes, filtered signal is carried out normalized, and it is calibrated, comprise eliminate baseline drift and (or) and the standard pulse wave compare correction etc., find out crest and the trough of signal in a wave period, obtain the ratio of trough value amplitude and crest value, utilize this ratio to estimate blood oxygen saturation (S _pO ₂), step 205:

S _pO ₂Computing formula can be S _pO ₂=kR/ (1+kR); Wherein R is the ratio of trough value and crest value, and k is one and adjusts operator, k＞1 and relevant with factors such as measurement posture, Measuring Time, temperature and skin complete.

Claims (4)

1. contactless automatic oxygen saturation measurement system, it is characterized in that, adopt video capture to carry out oxygen saturation measurement, this system comprises video acquisition module, primary color component separation module, ROI extraction module, time varying signal generation module, filtering and noise reduction module and signal calibration analysis module; Wherein:

Described video acquisition module is for the color video of the segment record human body wrist information of collection;

Described primary color component separation module is used for every two field picture of color video is carried out the primary colours separation, obtains wherein red component image;

Described ROI extraction module utilizes the texture information algorithm of wrist blood vessel and venation that above-mentioned every frame red component image is positioned, extract wherein comprise tremulous pulse in interior zone the ROI as this two field picture;

Described time varying signal generation module is used for getting the average of every all pixels of frame ROI image as the eigenvalue of this two field picture, generates the waveshape signal of tremulous pulse pulse amplitude change when time dependent;

Described filtering and noise reduction module is used for above-mentioned waveshape signal being carried out filtering and noise reduction, filtering arterial pulse wave wherein and the interference of respiratory wave;

Described signal calibration analysis module is used for filtered waveshape signal " x (n) " is carried out normalized, and its formula is:

$\hat{x}\left(n\right)=\frac{x\left(n\right)-x_{\min }\left(n\right)}{x_{\max }\left(n\right)-x_{\min }\left(n\right)},$

X wherein _Min(n) be the minima of signal x (n), x _Max(n) be the maximum of signal x (n); And it is calibrated, find out crest value and the trough value of signal in a wave period then, obtain the ratio of trough value and crest value, utilize this ratio to estimate blood oxygen saturation S _pO ₂Computing formula is:

S _pO ₂=kR/(1+kR)；

Wherein R is the ratio of trough value and crest value, and k is for adjusting operator.

2. contactless automatic oxygen saturation measurement as claimed in claim 1 system is characterized in that, if the color type of described color video is RGB, then directly video image is carried out primary colours and separates; If video image color type right and wrong RGB, then be translated into and carry out primary colours again behind the RGB color type and separate.

3. contactless automatic measurement method of blood oxygen saturation is characterized in that this method may further comprise the steps:

Step (1) is gathered the color video that comprises a segment record human body wrist information;

Step (2) is carried out primary colours to the colored every two field picture of gathering that frequently carries out and is separated, and extracts the red component image;

Step (3) is utilized by the texture location algorithm red component image that obtains is carried out the ROI location, extract comprise tremulous pulse in interior zone the ROI image as this two field picture, comprise the texture information of wrist blood vessel and venation in the described red component image;

Step (4) is got the average of every all pixels of frame ROI image as the eigenvalue of this two field picture, generates the waveshape signal of a time dependent time-amplitude of tremulous pulse pulse amplitude;

Step (5) is carried out filtering and noise reduction to waveshape signal, the interference of filtering arterial pulse wave and respiratory wave;

Step (6) is carried out normalization to filtered waveshape signal x (n), and its formula is:

$\hat{x}\left(n\right)=\frac{x\left(n\right)-x_{\min }\left(n\right)}{x_{\max }\left(n\right)-x_{\min }\left(n\right)},$

X wherein _Min(n) be the minima of signal x (n), x _Max(n) be the maximum of signal x (n); And it is calibrated, find out crest and the trough of signal in a wave period then, obtain the ratio of trough value and crest value, utilize this ratio to estimate blood oxygen saturation (S _pO ₂)

S _pO ₂=kR/(1+kR)；

Wherein R is the ratio of trough value and crest value, and k is one and adjusts operator.

4. contactless automatic measurement method of blood oxygen saturation as claimed in claim 1 is characterized in that, if the color type of described color video is RGB, then directly video image is carried out primary colours and separates; If video image color type right and wrong RGB, then be translated into and carry out primary colours again behind the RGB color type and separate.

Images