CN108720826A - Sport injury method for early warning based on laser speckle - Google Patents
Sport injury method for early warning based on laser speckle Download PDFInfo
- Publication number
- CN108720826A CN108720826A CN201810232251.0A CN201810232251A CN108720826A CN 108720826 A CN108720826 A CN 108720826A CN 201810232251 A CN201810232251 A CN 201810232251A CN 108720826 A CN108720826 A CN 108720826A
- Authority
- CN
- China
- Prior art keywords
- row
- value
- vector
- matrix
- speckle image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 208000027418 Wounds and injury Diseases 0.000 title claims abstract description 12
- 230000006378 damage Effects 0.000 title claims abstract description 12
- 208000014674 injury Diseases 0.000 title claims abstract description 12
- 230000000004 hemodynamic effect Effects 0.000 claims abstract description 14
- 230000000694 effects Effects 0.000 claims abstract description 11
- 210000004204 blood vessel Anatomy 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 119
- 239000011159 matrix material Substances 0.000 claims description 81
- 230000001186 cumulative effect Effects 0.000 claims description 31
- 239000008280 blood Substances 0.000 claims description 25
- 210000004369 blood Anatomy 0.000 claims description 24
- 210000003462 vein Anatomy 0.000 claims description 14
- 230000002792 vascular Effects 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 8
- 108090000623 proteins and genes Proteins 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000005314 correlation function Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 230000003287 optical effect Effects 0.000 claims description 4
- 239000004065 semiconductor Substances 0.000 claims description 4
- 239000006185 dispersion Substances 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 210000001519 tissue Anatomy 0.000 abstract description 11
- 238000004458 analytical method Methods 0.000 abstract description 8
- 238000012544 monitoring process Methods 0.000 abstract description 6
- 210000003205 muscle Anatomy 0.000 abstract description 4
- 238000004422 calculation algorithm Methods 0.000 abstract description 2
- 238000011156 evaluation Methods 0.000 abstract description 2
- 238000005259 measurement Methods 0.000 abstract description 2
- 230000003387 muscular Effects 0.000 abstract description 2
- 238000003384 imaging method Methods 0.000 description 11
- 238000012549 training Methods 0.000 description 5
- 230000017531 blood circulation Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 230000037396 body weight Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- 241000150786 Athletes Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002612 cardiopulmonary effect Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/026—Measuring blood flow
- A61B5/0261—Measuring blood flow using optical means, e.g. infrared light
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Cardiology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Hematology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Physiology (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of sport injury method for early warning based on laser speckle, shoots original speckle image, and carries out eliminating shake pretreatment；Using the speckle image under stationary state as benchmark initial value, variation caused by other uncertain factors such as noise in measurement process is reduced；Calculate speckle image contrasts value, and analysis is compared under a time series；Calculate the hemodynamics index of tested tissue, and early warning excessive amount of motion.Based on fast laser speckle image processing algorithm, the information correlativity obtained in a time series is strong, the hemodynamics index of local body tissue can effectively be monitored, objective evaluation sportsman part muscle group state, relative to existing motion monitoring tool, more focus on the blood vessel at local body position, muscular states monitoring.The advantages that this method has real-time good, and safety is good, and Evaluated effect is accurate, and mode of operation is simple.
Description
Technical field
The present invention relates to a kind of image procossing and application process, more particularly to a kind of sport injury based on laser speckle is pre-
Alarm method.
Background technology
During training, how to judge that the body index of sportsman has scientific arrangement training burden very high valence
Value.Unreasonable amount of exercise can make movement effects have a greatly reduced quality, and even generate injury to the health of sportsman.It is existing
MONITOR AND CONTROL SYSTEM and method have Usage data collection device acquisition movement sign data information to carry out correlation analysis again, such as：It is " a kind of
MONITOR AND CONTROL SYSTEM and method " (Chinese invention patent CN107670262A).But the monitor mode of contact can be to normal
Training is made troubles, and traditional monitor mode is mainly for the cardio-pulmonary function of whole body.
Laser speckle blood current imaging technology is emerging blood flow detection method in recent years, has non-contact, safety, quickly
The advantages that, the hemodynamics index of a certain section of vein blood vessel of body during the motion can be monitored, is trained for different muscle groups
Reference is provided.
The processing analysis of laser speckle image mainly has space to contrast analysis (Laser Speckle Spatial at present
Contrast Analysis, LSSCA) and time contrasts analysis (Laser Speckle Temporal Contrast
Analysis, LSTCA), such as：" a kind of laser speckle blood current imaging and analyzing method " (Chinese invention patent CN101485565),
" a kind of laser speckle blood stream imaging processing system and method " (Chinese invention patent CN102357033), a kind of " laser speckle blood
Flow imaging contrasts analysis method " (Chinese invention patent CN102429650) etc., but these above-mentioned methods are both for static right
As being detected, be not suitable for the monitoring of motion process.
Invention content
The problem of being applied to motion monitoring the present invention be directed to current laser speckle image, it is proposed that a kind of to be dissipated based on laser
The sport injury method for early warning of spot is based on fast laser speckle image processing algorithm, the information obtained in a time series
Correlation is strong, can effectively monitor the hemodynamics index of local body tissue, objective evaluation sportsman part muscle group state,
There is certain noise resisting ability, is used in motion detection field, it is helpful to scientific arrangement training athlete amount.
The technical scheme is that：A kind of sport injury method for early warning based on laser speckle, specifically includes following step
Suddenly：
1) test system building shoots original speckle image：Using semiconductor laser as light source, laser beam is through expanding
Mirror is uniformly impinged upon after expanding above measured's vein blood vessel to be measured, and image capturing system is the Near Infrared CCD for adding optical filter
Data are sent computer to be handled in real time by camera, CCD camera；
2) registration pretreatment is carried out to the original speckle image of acquisition, eliminate flating and carries out eliminating the pre- place of shake
Reason；
3) it calculates speckle image under time series and contrasts value；
4) obtained by step 3) under time series speckle image contrast value matrix be normalized with after pseudocolour picture processing
Point of interest is extracted, under the weights and subject's height and weight index that consider further that target blood, calculates tested tissue under time series
Hemodynamics index, calculate the variation ratio value of hemodynamics index therewith before not moved under stationary state；
5) according to the gender of measured, at the age, different threshold values, threshold value and variation ratio obtained by step 4) is arranged in movable basis
Example value is compared, and early warning is carried out to risk.
Step 2) the specific method：Convolution is carried out using the convolution kernel and original speckle image of a 3*3, it is every to obtain
The space criteria of a pixel is poor；It is iterated optimization using the image after normalized calculation of correlation function pair process of convolution；Most
Calculated registration parameter and cubic Bézier curves method complete original speckle image registration afterwards.
Step 3) the specific method：
3.1) picture size after pre-processing is m*n, by each pixel gray value deposit two-dimensional matrix F in image
(x, y) resettles a matrix Fsq(x, y) stores square of each pixel gray value in original speckle image, i.e. Fsq
Each value is square of numerical value in F (x, y) corresponding position in (x, y)；
Fsq(x, y)=F (x, y) * F (x, y) (1)
3.2) data of the preceding w rows of F (x, y) are done vector to add up, obtains the cumulative and vector S of an initializationrow, Srow
In each element correspond to w pixel in F (x, y) in same row addition value；It obtains in the same way at the beginning of one
The cumulative quadratic sum vector S of beginningizationrow_sq, Srow_sqIn each element homography FsqW member in (x, y) in same row
The value that element is added；
3.3) a new two-dimensional matrix N (x, y) is established, the cumulative and vector S that step 4 is generatedrowAs the matrix
The first row, that is, N (1,：)；N (1,：) plus w+1 row vectors F in original speckle image (w+1,：), then subtract original speckle pattern
As in the 1st row vector F (1,：) obtained result be exactly N (x, y) the second row N (2,：)；N (x, y) the second row N (2,：) plus
W+2 row vectors F in upper original speckle image (w+2,：), then subtract the 2nd row vector F in original speckle image (2,：) obtain
As a result be exactly N (x, y) the third line N (3,：)；The data of N (x, y) remaining row generate similar, the (i-1)-th row vectors of N (x, y)
In addition original speckle image w+i row vectors subtract the i-th row vector of original speckle image again can be obtained the i-th row vectors of N (x, y)；
N (i,：)=N (i-1,：)+F (w+i-1,：)-F (i-1,：) (2)
3.4) a new two-dimensional matrix N is establishedsq(x, y), square cumulative and vector S that third step is generatedrow_sqAs
The first row of the matrix；NsqThe data of (x, y) remaining row generate, N similar with previous stepsq(x, y) (i-1)-th row vector adds square
Battle array Fsq(x, y) w+i row vectors subtract matrix F againsqN can be obtained in (x, y) i-th row vectorsq(x, y) i-th row vector；
Nsq(i,：)=Nsq(i-1,：)+Fsq(w+i-1,：)-Fsq(i-1,：) (3)
3.5) the preceding w of N (x, the y) data arranged are done vector to add up, obtains the cumulative and vector S of an initializationcol, Scol
In the value that is added with w pixel in a line in each element homography N (x, y)；One is obtained in the same way
The cumulative quadratic sum vector S of a initializationcol_sq, Scol_sqIn each element homography NsqWith the w in a line in (x, y)
The value that a element is added；
3.6) a new two-dimensional matrix L (x, y) is established, the cumulative and vector S that previous step is generatedcolAs the matrix
First row, that is, L (：, 1)；L(：, 1) plus w+1 column vectors N in matrix N (x, y) (：, w+1), then subtract in matrix N (x, y)
1st column vector N (：, 1) obtained vector result be exactly L (x, y) secondary series L (：, 2)；The data of remaining row of L (x, y) generate
Similar, the (i-1)-th column vectors of L (x, y) subtract the i-th column vector of matrix N (x, y) again plus matrix N (x, y) w+i column vectors
Obtain the i-th column vectors of L (x, y)；
L(：, i)=L (：, i-1) ten N (：, ten i-1 of w)-N (：, i-1) and (4)
3.7) a new two-dimensional matrix L is establishedsq(x, y), square cumulative and vector S that the 6th step is generatedcol_sqAs
The first row of the matrix；LsqThe data of (x, y) remaining row generate, L similar with previous stepsq(x, y) (i-1)-th column vector adds square
Battle array Nsq(x, y) w+i column vectors subtract matrix N againsqL can be obtained in (x, y) i-th column vectorsq(x, y) i-th column vector；
Lsq(：, i) and=Lsq(：, i-1) and+Nsq(：, w+i-1) and-Nsq(：, i-1) and (5)
When until, the two-dimensional matrix that two sizes are (m-w+1) * (n-w+1) can be obtained, be L (x, y) and L respectivelysq
(x, y)；
3.8) it calculates and contrasts value
The original speckle image of m*n pixels is converted to (m-w+1) * (n-w+1) and contrasts value matrix C (x, y), contrasts value matrix
The value of middle certain point is：
The caliber and flow rate information that contrast value and contain blood vessel on vascular site.
The hemodynamics index calculating method of tested tissue under the step 4) time series：
A reference point (Nor_i, Nor_j) is chosen in contrasting value matrix, does not include tested biological tissue body usually
Background, be normalized with the value of contrasting that the reference point is all；
By 8 neighborhoods of reference point (Nor_i, Nor_j)【(Nor_i-1, Nor_j-1), (Nor_i-1, Nor_j), (Nor_
I-1, Nor_j+1), (Nor_i, Nor_j-1), (Nor_i, Nor_j+1), (Nor_i+1, Nor_j-1), (Nor_i+1, Nor_
J), (Nor_i+1, Nor_j+1)】Contrast value to carry out cumulative and find out mean value Cm；
The all values contrasted in value matrix C (x, y) are adjusted to newly to be worth by formula (7)
Wherein, A is Dynamic gene, is a constant, depends on the dispersion degree being respectively worth in C ' matrixes；
It shows that value matrix C ' (x, y) is contrasted in normalization using pseudocolour picture technology, changes Dynamic gene A and obtain best show
Show effect, there is very strong contrast in vascular tissue with background at this time, can extract out the profile of vein blood vessel；Use space cluster
Mode extracts target blood, draws rectangle ROI region；
Three points of observation are chosen in target blood, and being averaged for three points of observation is contrasted value and be denoted as L, vascular dynamics index
The formula of K is：
K=(α * L)/BMI (8)
Wherein, α is the weights that target blood is obtained according to importance, in the range of [0-1]；BMI is subject's height body
Weight index, unit kg/m2。
The beneficial effects of the present invention are：The present invention is based on the sport injury method for early warning of laser speckle, relative to existing
Motion monitoring tool, more focus on the blood vessel at local body position, muscular states monitoring.This method has real-time good, peace
The advantages that good perfection, Evaluated effect is accurate, and mode of operation is simple.
Description of the drawings
Fig. 1 is that the present invention is based on the excessive amount of motion method for early warning flow charts of laser speckle blood current imaging technology；
Fig. 2 is imaging system pictorial diagram of the present invention；
Fig. 3 is elimination shake effect comparison schematic diagram after image registration of the present invention；
Fig. 4 is initial time imaging effect figure of the present invention；
Imaging effect figure when Fig. 5 is excessive movement of the present invention.
Specific implementation mode
A kind of excessive amount of motion method for early warning based on laser speckle blood current imaging technology, is broadly divided into 4 parts：First
It is the original speckle image of shooting to divide, and carries out eliminating shake pretreatment；Second part is to make the speckle image under stationary state
On the basis of initial value, reduce variation caused by other uncertain factors such as noise in measurement process；Part III is to calculate speckle
Image contrasts value, and analysis is compared under a time series；Part IV be calculate tested tissue blood it is dynamic
Power index, and early warning excessive amount of motion.
The present invention hardware device include：Semiconductor laser, beam expanding lens, Near Infrared CCD camera, optical filter.
Excessive amount of motion method for early warning flow chart based on laser speckle blood current imaging technology as shown in Figure 1, skill of the invention
Art scheme specifically comprises the following steps：
The first step：Test system building as shown in Figure 2 shoots original speckle image：Using semiconductor laser as light source,
Laser beam is uniformly impinged upon after beam expanding lens expands above measured's vein blood vessel to be measured, and image capturing system is to add optical filtering
The Near Infrared CCD camera of piece, can be with isolation environment light pollution.Camera is calculated by RJ45 interfaces using the connection of class ethernet netting twine
Machine is handled in real time.
When being monitored to body part tissue, region of interest should be kept static when finding a view.
Second step：Original speckle image is acquired using laser speckle blood current imaging device, and carries out registration pretreatment and eliminates
Flating.
Convolution is carried out using the convolution kernel and original speckle image of a 3*3, it is poor with the space criteria for obtaining each pixel.
It is iterated optimization using the image after normalized calculation of correlation function pair process of convolution.Last calculated registration
Parameter and cubic Bézier curves method complete original speckle image registration.The effect for eliminating shake is shown in attached drawing 3, and the left side is in figure
Before processing, the right is to scheme after handling.
Third walks：Picture size after pretreatment is m*n, each pixel gray value in image is stored in Two-Dimensional Moment
Battle array F (x, y).Resettle a matrix Fsq(x, y) stores each in original speckle imageSquare of pixel(should be " as
Square of vegetarian refreshments gray value "), i.e. FsqEach value is square of numerical value in F (x, y) corresponding position in (x, y).
Fsq(x, y)=F (x, y) * F (x, y) (formula 1)
4th step：The data of the preceding w rows of F (x, y) are done vector to add up, obtain the cumulative and vector S of an initializationrow,
SrowIn each element correspond to w pixel in F (x, y) in same row addition value；One is obtained in the same way
The cumulative quadratic sum vector S of initializationrow_sq, Srow_sqIn each element homography FsqW in (x, y) in same row
The value that element is added.
5th step：A new two-dimensional matrix N (x, y) is established, the cumulative and vector S that step 4 is generatedrowAs the square
Battle array the first row, that is, N (1,：)；N (1,：) plus w+1 row vectors F in original speckle image (w+1,：), then subtract original speckle
1st row vector F in image (1,：) obtained result be exactly N (x, y) the second row N (2,：)；N (x, y) the second row N (2,：)
In addition w+2 row vectors F in original speckle image (w+2,：), then subtract the 2nd row vector F in original speckle image (2,：) obtain
Result be exactly N (x, y) the third line N (3,：)；The data of N (x, y) remaining row generate it is similar, the (i-1)-th rows of N (x, y) to
Amount plus original speckle image w+i row vectors subtract again the i-th row vector of original speckle image can be obtained the i-th rows of N (x, y) to
Amount；
N (i,：)=N (i-1,：)+F (w+i-1,：)-F (i-1,：) (formula 2)
6th step：Establish a new two-dimensional matrix Nsq(x, y), square cumulative and vector S that third step is generatedrow_sq
The first row as the matrix；NsqThe data of (x, y) remaining row generate, N similar with previous stepsq(x, y) (i-1)-th row vector adds
Upper matrix Fsq(x, y) w+i row vectors subtract matrix F againsqN can be obtained in (x, y) i-th row vectorsq(x, y) i-th row vector；
Nsq(i,：)=Nsq(i-1,：)+Fsq(w+i-1,：)-Fsq(i-1,：) (formula 3)
7th step：The preceding w of N (x, the y) data arranged are done vector to add up, obtain the cumulative and vector S of an initializationcol,
ScolIn the value that is added with w pixel in a line in each element homography N (x, y)；It obtains in the same way
The cumulative quadratic sum vector S of one initializationcol_sq, Scol_sqIn each element homography NsqWith in a line in (x, y)
W element be added value.
8th step：A new two-dimensional matrix L (x, y) is established, the cumulative and vector S that previous step is generatedcolAs the square
Battle array first row, that is, L (：, 1)；L(：, 1) plus w+1 column vectors N in matrix N (x, y) (：, w+1), then subtract matrix N (x, y)
In the 1st column vector N (：, 1) obtained vector result be exactly L (x, y) secondary series L (：, 2).The data life of remaining row of L (x, y)
At similar, the (i-1)-th column vectors of L (x, y) plus matrix N (x, y) w+i column vectors subtract again matrix N (x, y) i-th arrange to
Measure the i-th column vectors of L (x, y)；
L(：, i)=L (：, i-1)+N (：, ten i-1 of w)-N (：, i-1) and (formula 4)
9th step：Establish a new two-dimensional matrix Lsq(x, y), square cumulative and vector S that the 6th step is generatedcol_sq
First row as the matrix；LsqThe data of (x, y) remaining row generate, L similar with previous stepsq(x, y) (i-1)-th column vector adds
Upper matrix Nsq(x, y) w+i column vectors subtract matrix N againsqL can be obtained in (x, y) i-th column vectorsq(x, y) i-th column vector；
Lsq(：, i) and=Lsq(：, i-1) and+Nsq(：, w+i-1) and-Nsq(：, i-1) and (formula 5)
When until the step, the two-dimensional matrix that two sizes are (m-w+1) * (n-w+1) can be obtained, be L (x, y) respectively
And Lsq(x, y).
Tenth step：Value is contrasted in calculating
The original speckle image of m*n pixels can be converted to (m-w+1) * (n-w+1) and contrast value matrix C (x, y)
The value for contrasting certain point in value matrix is：
The caliber and flow rate information that contrast value and contain blood vessel on vascular site.
11st step：A reference point (Nor_i, Nor_j) is chosen in contrasting value matrix, does not include tested life usually
The background of object organizer is normalized with the value of contrasting that the reference point is all.
By 8 neighborhoods of reference point (Nor_i, Nor_j)【(Nor_i-1, Nor_j-1), (Nor_i-1, Nor_j), (Nor_
I-1, Nor_j+1), (Nor_i, Nor_j-1), (Nor_j, Nor_j+1), (Nor_i+1, Nor_j-1), (Nor_i+1, Nor_
J), (Nor_i+1, Nor_j+1)】Contrast value to carry out cumulative and find out mean value Cm。
The all values contrasted in value matrix C (x, y) are adjusted to newly to be worth by formula (7)
Wherein, A is Dynamic gene, is a constant, depends on the dispersion degree being respectively worth in C ' matrixes.
12nd step：Contrast value matrix C ' (x, y) using the display normalization of pseudocolour picture technology, changes Dynamic gene A and obtain
Display effect most preferably is obtained, there is very strong contrast in vascular tissue with background at this time, can extract initial time arm elbow just
The profile of medium sized vein blood vessel.The mode of use space cluster extracts target blood, draws rectangle ROI region.
13rd step：Three points of observation are chosen in target blood, and being averaged for three points of observation is contrasted value and be denoted as L, blood
The formula of pipe dynamic index K is：
K=(α * L)/BMI (formula 8)
Wherein, α is the weights that target blood is obtained according to importance, in the range of [0-1].BMI is subject's height body
Weight index, unit kg/m2.Static index before subject does not move is denoted as K0。
14th step：Pseudocolour picture under all time serieses is obtained, and is calculated under different time, same target blood
Hemodynamics index K, consider index Kt and initial time index K under the t times0The variation Z to comparet
Zt=| Kt-K0|/K0* 100% (formula 9)
ZtNumerical value is bigger, illustrates that blood vessel state variation is more violent.According to the gender of subject, age, movable basis setting
Different threshold value Zx, work as Zt> Zx, system can alarm, and subject is reminded to have the risk of excessive movement.
In bicipital muscle of arm training, carries out movement and endanger early warning, the body part of shooting is arm median basilic vein blood vessel.
Referring to Fig.1, the laser wavelength of selection is 785nm, and 785nm bandpass filters are installed additional before the camera lens of CCD camera.
After the laser that laser is sent out is expanded by laser beam expander, uniform irradiation on tested vein blood vessel,
Back scattering forms speckle on CCD camera surface, and CCD camera is imaged speckle, and original speckle image is sent to calculating
Machine is handled.
It is to the concrete mode of excessive amount of motion progress early warning in the present invention：
The first step：The arm median basilic vein image under stationary state is shot, original laser speckle image is obtained.Use one
The convolution kernel of a 3*3 and original speckle image carry out convolution, using the image after normalized calculation of correlation function pair convolution into
Row iteration optimizes, and obtains reusing the speckle image after cubic Bézier curves method is registrated after registration parameter.The image
Resolution ratio be 1040*1392.
Second step：With the gray value of each pixel in two-dimensional matrix F (x, y) the storage images of 1040*1392 sizes；
With the two-dimensional matrix F of 1040*1392 sizessq(x, y) stores square of the gray value of each pixel in image；Select 7*7
Dimensional slip window.
Third walks：The data of preceding 7 row of F (x, y) are done vector to add up, obtain the cumulative and vector S of an initializationrow；
By FsqThe data of preceding 7 row of (x, y) do vector and add up, and obtain the cumulative and vector S of an initializationrow_sq。
Third walks：The two-dimensional matrix N (x, y) for establishing a 1034*1392 size, by SrowAs N (1,：)；N (x, y)
I-1 row vectors subtract the i-th row vectors of F (x, y) plus F (x, y) w+i row vectors and the i-th row vectors of N (x, y) can be obtained again.
4th step：Establish the two-dimensional matrix Ns of a 1034*1392 sizeq(x, y), by Srow_sqAs Nsq(1,：)；Nsq
(x, y) (i-1)-th row vector adds Fsq(x, y) w+i row vectors subtract F againsqN can be obtained in (x, y) i-th row vectorsq(x, y) i-th
Row vector.
5th step：The data of preceding 7 row of N (x, y) are done vector to add up, obtain the cumulative and vector S of an initializationcol；
By NsqThe data of preceding 7 row of (x, y) do vector and add up, and obtain the cumulative and vector S of an initializationcol_sq。
6th step：The two-dimensional matrix L (x, y) for establishing a 1034*1386 size, by ScolAs L (：, 1)；L (x, y)
I-1 column vectors plus matrix N (x, y) w+i column vectors subtract again the i-th column vector of matrix N (x, y) obtain L (x, y) i-th arrange to
Amount.
7th step：Establish the two-dimensional matrix L of a 1034*1386 sizesq(x, y), by Scol_sqAs Lsq(1,：)；Lsq
(x, y) (i-1)-th column vector adds matrix Nsq(x, y) w+i column vectors subtract matrix N againsq(x, y) i-th column vector obtains Lsq
(x, y) i-th column vector.
8th step：Contrast value according to the calculating of formula 6, result is stored in the Matrix C (x, y) of 1034*1386 sizes.
9th step：Carry out the space cluster analysis based on pixel
In all time-series images, if certain moment Two-Dimensional Speckle image is represented by two-dimensional matrix C (x, y, t)
CI, j(1≤i≤M, 1≤j≤N) is a bit on t moment image, and image size is M*N, with base value CKTo CI, jInto
Row normalized, normalized mode are
Shake effect comparison schematic diagram is eliminated after image registration as shown in Figure 3 in this example, cross in Fig. 3 is selected to sit
The mean value in 8 fields is as base value c around point in markK, at this time it is considered that all time serieses in picture all in time
Set up correlation.The left side is the image that hand shake has artifact, and the right is the image of elimination artifact after registration operation.
Tenth step：The median basilic vein of subject is shot in time series t, is shown with Pseudo-color Technique each
Value matrix is contrasted in two dimension normalization on a time point, and pseudo color image at this time is exactly blood flow velocity distributed image, different
Color depth reflects flow velocity speed, and flow velocity is faster, and color is deeper.
In the flow velocity figure, blood vessel has very strong comparison with background, can use the mode of image clustering, draw ROI region,
Subject is measured in t0The hemodynamics index K of the median basilic vein blood vessel at moment0。
12nd step：During Athletess, ROI region is drawn to the original speckle image that different moments obtain
Afterwards, in different times in obtain the hemodynamics index K of same target bloodt, continuously detect ZtValue, works as ZtWhen more than threshold value
It alarms, value pseudocolour picture is contrasted in specifying information such as Fig. 4,5 when being initial time and excessive movement respectively；Arm blood
Pipe is expanded during the motion, and flow velocity becomes faster.Black surround indicates that target blood median basilic vein, three mark points are to measure to serve as a contrast
The pixel of ratio.The cross frame on the right sides Fig. 4 is reference point.Hemodynamics index is as K using in Fig. 40, set the threshold of excessive movement
Value is 35%, when sportsman largely moves to its limit, ZtValue is 35.27%, and system is alarmed, and illustrates that this system has
Practicability.
The present invention is contrasted value and is measured the variation of vascular flow rate and blood vessels caliber using normalized, analyzes bodily tissue
The hemodynamics index of vein blood vessel is to determine whether excessive movement, is the standard quantified completely, will not bring subjective judgement into.This
Invention arithmetic speed is fast, can realize the real-time display of pseudocolour picture, meets the requirement of practical application.The present invention can reject original
The several special pixel pair graphs of certain in beginning speckle image are influenced as caused by, there is certain anti-noise function.
Claims (4)
1. a kind of sport injury method for early warning based on laser speckle, which is characterized in that specifically comprise the following steps：
1) test system building shoots original speckle image：Using semiconductor laser as light source, laser beam expands through beam expanding lens
It uniformly being impinged upon after beam above measured's vein blood vessel to be measured, image capturing system is the Near Infrared CCD camera for adding optical filter,
Data are sent computer to be handled in real time by CCD camera；
2) registration pretreatment is carried out to the original speckle image of acquisition, eliminate flating and carries out eliminating shake pretreatment；
3) it calculates speckle image under time series and contrasts value matrix；
4) value matrix that contrasts of speckle image is normalized and is extracted with after pseudocolour picture processing under time series obtained by step 3)
Point of interest under the weights and subject's height and weight index that consider further that target blood, calculates the blood of tested tissue under time series
Hydraulic power index calculates the variation ratio value of hemodynamics index therewith before not moved under stationary state；
5) according to the gender of measured, at the age, different threshold values, threshold value and variation ratio value obtained by step 4) is arranged in movable basis
It is compared, early warning is carried out to risk.
2. the sport injury method for early warning based on laser speckle according to claim 1, which is characterized in that the step 2) tool
Body method：Convolution is carried out using the convolution kernel and original speckle image of a 3*3, it is poor with the space criteria for obtaining each pixel；
It is iterated optimization using the image after normalized calculation of correlation function pair process of convolution；Last calculated registration
Parameter and cubic Bézier curves method complete original speckle image registration.
3. the sport injury method for early warning according to claim 1 or claim 2 based on laser speckle, which is characterized in that the step
3) specific method：
3.1) picture size after pre-processing is m*n, by each pixel gray value deposit two-dimensional matrix F in image (x,
Y), a matrix F is resettledsq(x, y) stores square of each pixel gray value in original speckle image, i.e. Fsq(x,
Y) each value is square of numerical value in F (x, y) corresponding position in；
Fsq(x, y)=F (x, y) * F (x, y) (1)
3.2) data of the preceding w rows of F (x, y) are done vector to add up, obtains the cumulative and vector S of an initializationrow, SrowIn it is every
One element all corresponds to the value of the addition of w pixel in F (x, y) in same row；An initialization is obtained in the same way
Cumulative quadratic sum vector Srow_sq, Srow_sqIn each element homography FsqW element phase in (x, y) in same row
The value added；
3.3) a new two-dimensional matrix N (x, y) is established, the cumulative and vector S that step 4 is generatedrowAs the matrix
A line, that is, N (1,：)；N (1,：) plus w+1 row vectors F in original speckle image (w+1,：), then subtract in original speckle image
1st row vector F (1,：) obtained result be exactly N (x, y) the second row N (2,：)；N (x, y) the second row N (2,：) plus original
W+2 row vectors F in beginning speckle image (w+2,：), then subtract the 2nd row vector F in original speckle image (2,：) obtained result
Be exactly N (x, y) the third line N (3,：)；The data generation of N (x, y) remaining row is similar, and the (i-1)-th row vectors of N (x, y) add
Original speckle image w+i row vectors subtract the i-th row vector of original speckle image and the i-th row vectors of N (x, y) can be obtained again；
N (i,：)=N (i-1,：)+F (w+i-1,：)-F (i-1,：) (2)
3.4) a new two-dimensional matrix N is establishedsq(x, y), square cumulative and vector S that third step is generatedrow_sqAs the square
The first row of battle array；NsqThe data of (x, y) remaining row generate, N similar with previous stepsq(x, y) (i-1)-th row vector adds matrix Fsq
(x, y) w+i row vectors subtract matrix F againsqN can be obtained in (x, y) i-th row vectorsq(x, y) i-th row vector；
Nsq(i,：)=Nsq(i-1,：)+Fsq(w+i-1,：)-Fsq(i-1,：) (3)
3.5) the preceding w of N (x, the y) data arranged are done vector to add up, obtains the cumulative and vector S of an initializationcol, ScolIn it is every
The value being added with w pixel in a line in one element all homography N (x, y)；It obtains in the same way at the beginning of one
The cumulative quadratic sum vector S of beginningizationcol_sq, Scol_sqIn each element homography NsqWith w member in a line in (x, y)
The value that element is added；
3.6) a new two-dimensional matrix L (x, y) is established, the cumulative and vector S that previous step is generatedcolAs the matrix
One row i.e. L (：, 1)；L(：, 1) plus w+1 column vectors N in matrix N (x, y) (：, w+1), then subtract the 1st in matrix N (x, y)
Column vector N (：, 1) obtained vector result be exactly L (x, y) secondary series L (：, 2)；The data of remaining row of L (x, y) generate and this
Similar, the (i-1)-th column vectors of L (x, y) subtract the i-th column vector of matrix N (x, y) plus matrix N (x, y) w+i column vectors and obtain L again
(x, y) i-th column vector；
L(：, i)=L (：, i-1) ten N (：, ten i-1 of w)-N (：, i-1) and (4)
3.7) a new two-dimensional matrix L is establishedsq(x, y), square cumulative and vector S that the 6th step is generatedcol_sqAs the square
The first row of battle array；LsqThe data of (x, y) remaining row generate, L similar with previous stepsq(x, y) (i-1)-th column vector adds matrix Nsq
(x, y) w+i column vectors subtract matrix N againsqL can be obtained in (x, y) i-th column vectorsq(x, y) i-th column vector；
Lsq(：, i) and=Lsq(：, i-1) and+Nsq(：, w+i-1) and-Nsq(：, i-1) and (5)
When until, the two-dimensional matrix that two sizes are (m-w+1) * (n-w+1) can be obtained, be L (x, y) and L respectivelysq(x,
y)；
3.8) it calculates and contrasts value
The original speckle image of m*n pixels is converted to (m-w+1) * (n-w+1) and contrasts value matrix C (x, y), contrasts certain in value matrix
The value of any is：
The caliber and flow rate information that contrast value and contain blood vessel on vascular site.
4. the sport injury method for early warning based on laser speckle according to claim 3, which is characterized in that when the step 4)
Between under sequence tested tissue hemodynamics index calculating method：
A reference point (Nor_i, Nor_j) is chosen in contrasting value matrix, does not include the back of the body of tested biological tissue body usually
Scape is normalized with the value of contrasting that the reference point is all；
By 8 neighborhoods of reference point (Nor_i, Nor_j)【(Nor_i-1, Nor_j-1), (Nor_i-1, Nor_j), (Nor_i-1,
Nor_j+1), (Nor_i, Nor_j-1), (Nor_i, Nor_j+1), (Nor_i+1, Nor_j-1), (Nor_i+1, Nor_j),
(Nor_i+1, Nor_j+1)】Contrast value to carry out cumulative and find out mean value cm；
The all values contrasted in value matrix C (x, y) are adjusted to newly to be worth by formula (7)
Wherein, A is Dynamic gene, is a constant, depends on the dispersion degree being respectively worth in C ' matrixes；
Contrast value matrix C ' (x, y) using the display normalization of pseudocolour picture technology, changes Dynamic gene A and obtain best display effect
Fruit, at this time vascular tissue and background have very strong contrast, can extract out the profile of vein blood vessel；The mode of use space cluster
Target blood is extracted, rectangle ROI region is drawn；
Three points of observation are chosen in target blood, and being averaged for three points of observation is contrasted value and be denoted as L, vascular dynamics index K's
Formula is：
K=(α * L)/BMI (8)
Wherein, α is the weights that target blood is obtained according to importance, in the range of [0-1]；BMI is that subject's height and weight refers to
Mark, unit kg/m2。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810232251.0A CN108720826A (en) | 2018-03-20 | 2018-03-20 | Sport injury method for early warning based on laser speckle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810232251.0A CN108720826A (en) | 2018-03-20 | 2018-03-20 | Sport injury method for early warning based on laser speckle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108720826A true CN108720826A (en) | 2018-11-02 |
Family
ID=63940959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810232251.0A Pending CN108720826A (en) | 2018-03-20 | 2018-03-20 | Sport injury method for early warning based on laser speckle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108720826A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112890793A (en) * | 2019-12-04 | 2021-06-04 | 德尔格制造股份两合公司 | Apparatus and method for presenting medical alerts |
| CN113303773A (en) * | 2021-05-20 | 2021-08-27 | 武汉理工大学 | Motion risk assessment method and device and readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101485565A (en) * | 2009-02-13 | 2009-07-22 | 华中科技大学 | Laser speckle blood current imaging and analyzing method |
| CN101744622A (en) * | 2009-12-31 | 2010-06-23 | 上海量科电子科技有限公司 | Amount of exercise measurement method and system thereof |
| US20120071769A1 (en) * | 2009-02-17 | 2012-03-22 | Andrew Dunn | Methods of producing laser speckle contrast images |
| CN103330557A (en) * | 2013-06-25 | 2013-10-02 | 上海理工大学 | Exposure time determination-based laser speckle blood flow imaging method |
| CN107389680A (en) * | 2017-06-29 | 2017-11-24 | 华中科技大学鄂州工业技术研究院 | A kind of quantitative viscoplasticity detection method |
-
2018
- 2018-03-20 CN CN201810232251.0A patent/CN108720826A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101485565A (en) * | 2009-02-13 | 2009-07-22 | 华中科技大学 | Laser speckle blood current imaging and analyzing method |
| US20120071769A1 (en) * | 2009-02-17 | 2012-03-22 | Andrew Dunn | Methods of producing laser speckle contrast images |
| CN101744622A (en) * | 2009-12-31 | 2010-06-23 | 上海量科电子科技有限公司 | Amount of exercise measurement method and system thereof |
| CN103330557A (en) * | 2013-06-25 | 2013-10-02 | 上海理工大学 | Exposure time determination-based laser speckle blood flow imaging method |
| CN107389680A (en) * | 2017-06-29 | 2017-11-24 | 华中科技大学鄂州工业技术研究院 | A kind of quantitative viscoplasticity detection method |
Non-Patent Citations (1)
| Title |
|---|
| 贾亚威: "激光散斑血流成像对中医理疗功效的检测", 《光学精密工程》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112890793A (en) * | 2019-12-04 | 2021-06-04 | 德尔格制造股份两合公司 | Apparatus and method for presenting medical alerts |
| CN112890793B (en) * | 2019-12-04 | 2024-03-15 | 德尔格制造股份两合公司 | Apparatus and method for presenting medical alerts |
| CN113303773A (en) * | 2021-05-20 | 2021-08-27 | 武汉理工大学 | Motion risk assessment method and device and readable storage medium |
| CN113303773B (en) * | 2021-05-20 | 2023-02-28 | 武汉理工大学 | Motion risk assessment method and device and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3405105B1 (en) | Method and apparatus for estimating heart rate | |
| Briers | Laser speckle contrast imaging for measuring blood flow. | |
| US9380935B2 (en) | Image processing apparatus, image processing method, and program | |
| KR102014104B1 (en) | Ultrasound examination system and ultrasound examination method | |
| CN104809480B (en) | A kind of eye fundus image Segmentation Method of Retinal Blood Vessels based on post-class processing and AdaBoost | |
| CN113408508B (en) | Transformer-based non-contact heart rate measurement method | |
| CN111111111A (en) | Real-time fitness monitoring system and method | |
| Datcu et al. | Noncontact automatic heart rate analysis in visible spectrum by specific face regions | |
| WO2019046003A1 (en) | Speckle contrast analysis using machine learning for visualizing flow | |
| JP2010532699A (en) | Laser speckle imaging system and method | |
| CN107485383B (en) | Speckle blood flow imaging method and device based on component analysis | |
| CN106419890B (en) | Blood flow velocity measuring device and method based on space-time modulation | |
| CN109009052A (en) | The embedded heart rate measurement system and its measurement method of view-based access control model | |
| CN108720826A (en) | Sport injury method for early warning based on laser speckle | |
| Przybyło | A deep learning approach for remote heart rate estimation | |
| JP7183590B2 (en) | Ophthalmic image processing device, OCT device, and ophthalmic image processing program | |
| US10803601B2 (en) | Rapid assessment and visual reporting of local particle velocity | |
| Wei et al. | Center of mass estimation for balance evaluation using convolutional neural networks | |
| JP7147888B2 (en) | Ophthalmic image processing device, OCT device, and ophthalmic image processing program | |
| EP2693397A1 (en) | Method and apparatus for noise reduction in an imaging system | |
| Marin et al. | Numerical observer for cardiac motion assessment using machine learning | |
| Niessen et al. | In vivo analysis of trabecular bone architecture | |
| CN106446810A (en) | Computer vision method used for mental state analysis | |
| Khalil et al. | Microvascular blood flow with laser speckle contrast imaging: analysis of static scatterers effect through modelling and simulation | |
| JP2006333902A5 (en) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181102 |
|
| RJ01 | Rejection of invention patent application after publication |