CN107749049A - A kind of vein distribution display method and device - Google Patents
A kind of vein distribution display method and device Download PDFInfo
- Publication number
- CN107749049A CN107749049A CN201710804444.4A CN201710804444A CN107749049A CN 107749049 A CN107749049 A CN 107749049A CN 201710804444 A CN201710804444 A CN 201710804444A CN 107749049 A CN107749049 A CN 107749049A
- Authority
- CN
- China
- Prior art keywords
- image
- chain
- mrow
- blood vessel
- point
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 210000003462 vein Anatomy 0.000 title claims abstract description 37
- 210000004204 blood vessel Anatomy 0.000 claims abstract description 87
- 239000008280 blood Substances 0.000 claims abstract description 19
- 210000004369 blood Anatomy 0.000 claims abstract description 19
- 238000012545 processing Methods 0.000 claims abstract description 14
- 230000002708 enhancing effect Effects 0.000 claims description 12
- 239000000203 mixture Substances 0.000 claims description 5
- 230000014509 gene expression Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 claims description 2
- 239000003086 colorant Substances 0.000 abstract description 3
- 210000002615 epidermis Anatomy 0.000 abstract description 3
- 210000003491 skin Anatomy 0.000 description 13
- 238000001914 filtration Methods 0.000 description 6
- 238000004040 coloring Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000007670 refining Methods 0.000 description 2
- 230000002792 vascular Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000002845 discoloration Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/40—Filling a planar surface by adding surface attributes, e.g. colour or texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a kind of vein distribution display method and device, wherein, methods described includes：Gather the infrared image and visible images at position to be detected；The infrared image is pre-processed, obtains intermediate image；The curvature maximum of the intermediate image is asked for, and filters out noise spot, to obtain feature dot image；Based on blood vessel bearing of trend, processing is attached to the blood vessel characteristic point in the feature dot image, obtains vessel branch image；Vessel branch in the vessel branch image is connected into vein network, to form the center line of human vas, and removes the burr on the center line, the center line image after being handled；The target blood region in the center line image is determined, and corresponding with the target blood region in the visible images is partially filled with blood vessel color.The technical scheme that the application provides, the problem of can solve the problem that the missing image shown by traditional vein display system or epidermis primary colors can not be recovered completely.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of vein distribution display method and device.
Background technology
Vein distribution carries out enhancing to venae subcutaneae blood vessel using the means of infrared imaging and shown.This technology can be notable
Improve the success rate punctured.And the display effect of infrared image lost the realistic colour of skin, therefore to tie similar to gray-scale map
Close the distribution of the vein in the colouring information and infrared image of visible images to blend, to improve display effect.So occur as soon as
One it is new the problem of：The color of hand skin, and skin can not be fully retained with infrared image integration technology for conventional visible ray
Contrast with blood vessel is still not obvious enough.Therefore, it is necessary to from infrared image, position, width information blood vessel extract,
Individually it is superimposed upon on visible images, to reach the purpose of enhancing display.
Currently, the method for extracting infrared image upper vein blood vessel mainly has three classes.First kind method is first to original image
Carry out smothing filtering and remove noise, then directly choose area-of-interest and carry out thresholding processing.This kind of method is to image matter
Amount requires higher, or needs given area-of-interest, and otherwise global threshold can not accurately extract Major Vessels, also suffer from because
The interference for the noise that uneven illumination is brought.Some improved thresholding methods can improve this situation, but still can not be compared with
Good elimination noise.The method that second class method uses vascular pattern matching.Gauss matched filtering device can be used to carry out secondary
Filtering obtains refining vascular skeleton, and this method can not determine blood vessel width, and noise is still more.In addition, the existing skill also having
Art using non-down-sampled Wavelet Transform extraction blood vessel feature, reuses segmentation Spline Method and matched, this method is higher first
This matched filter is preferred, but needs the width of estimation blood vessel in the picture in advance.3rd class is come really by image curvature
Determine the method for vessel position：The curvature of image is calculated line by line, finds local maximum to determine blood vessel center line position, is a kind of
Relatively stable extracting method.This method is also used for the preliminary extraction of vessel position, but is still required for refining and is made an uproar to remove
Sound.
The content of the invention
In order to solve problem of the prior art, the embodiments of the invention provide a kind of vein distribution display method and device.
The technical scheme is as follows：
On the one hand, a kind of vein distribution display method, methods described include：
Gather the infrared image and visible images at position to be detected；
The infrared image is pre-processed, to remove noise and strengthen contrast, obtains intermediate image；
The curvature maximum of the intermediate image is asked for, and filters out noise spot, to obtain comprising all blood vessel characteristic points
Feature dot image, wherein, the result images include blood vessel bearing of trend corresponding to each blood vessel characteristic point；
Based on the blood vessel bearing of trend, processing is attached to the blood vessel characteristic point in the feature dot image, obtained
Vessel branch image；
Vessel branch in the vessel branch image is connected into vein network, to form the center line of human vas,
And the burr on the center line is removed, the center line image after being handled；
Determine the target blood region in the center line image, and in the visible images with the target blood
Blood vessel color is partially filled with corresponding to region.
Further, after the infrared image and visible images at position to be detected is gathered, methods described also includes：
The infrared image and the visible images are subjected to registration, so that the position of two images is corresponding.
Further, being attached processing to the blood vessel characteristic point in the feature dot image includes：
First then connection corresponds at a distance of two blood vessel characteristic points of the vacancy of only one pixel according to blood vessel characteristic point
Blood vessel bearing of trend connection interruption vessel centerline, the blood vessel characteristic point scattered is linked to be vessel branch.
Further, it is determined that the target blood region in the center line image includes：
The two-sided search curvature of the bearing of trend of blood vessel characteristic point is more than zero scope on heart line in the blood vessel, and will search
Width of the obtained range wide as blood vessel at this, and calculate the mean breadth of blood vessel；
Expansive working is carried out to the center line image with the mean breadth, obtains target blood region.
Further, the intermediate image determines in the following manner：
The Otsu threshold of the infrared image is calculated, and human body skin in the infrared image is extracted using Otsu threshold
Region；
Contrast enhancing is carried out to the region of the human body skin according to the following equation：
Wherein, nom1 (x, y) represents pixel value corresponding to the region of the enhanced human body skin of contrast, I_ROI (x, y)
Pixel value corresponding to the region of human body skin before expression contrast enhancing, min expressions take minimum operation, and max represents to take
Maximum operation, (x, y) represent the coordinate of pixel；
Pixel value corresponding to the region of the enhanced human body skin of the contrast is filtered according to the following equation：
F (x, y)=nom1 (x, y) * GaussianKernel (11,2.5)；
Wherein, GaussianKernel is Gaussian kernel, and F (x, y) is filtered pixel value；
Carry out contrast enhancing operation again to the filtered pixel value according to the following equation：
As F (x, y) ＜ mean,
As F (x, y) ＞ mean,
Wherein, mean is pixel average, and V is the variance of pixel value, MsetFor default output image average, VsetTo be pre-
If output image variance, I1(x, y) represents to carry out the enhanced pixel value of contrast again；
Using the image for carrying out the enhanced pixel value composition of contrast again as the intermediate image.
Further, the Otsu threshold determines in the following manner：
The histogram of image to be split is provided, a threshold value t is calculated in traversal so that side between the class of threshold value t both sides
It is poor maximum, i.e.,：
Pixnum (I | I ＞ t) (ave (I | I ＞ t)-ave (I))2+ pixnum (I | I ＜ t) (ave (I | I ＜ t)-
ave(I))2When obtaining maximum, the value of the threshold value t is determined；
Wherein, pixnum is the pixel count of gray value within the specified range, and ave is pixel average, and I represents described infrared
The pixel value of image.
Further, being attached processing to the blood vessel characteristic point in the feature dot image includes：
All vessel segments in the feature dot image are extracted, the connected blood vessels characteristic point of all composition 8- syntoples can be regarded as one
Bar point chain；
All point chains are extracted, a chain is numbered；Wherein, described chain characterizes vessel centerline；
Count all point chains pair for being only separated by a pixel；
In the point chain pair that most nearby connection statistics go out, the bearing of trend of tie point is the blood vessel characteristic point bearing of trend of both sides
The average value of angle.
Further, methods described also includes：
Having a chain and numbering in the feature dot image is extracted, the end points for finding a chain is marked；
Long point chain of the length in pixels more than 8 is found out from described chain, and writes down the long point chain numbering；
Other chains in every long chain periphery distance to a declared goal are found, and the numbering of the point chain searched out is attached to the length
After point chain numbering.
Further, the center line image after determination is handled in the following manner：
Take principal point chain most long in vessel branch image, and find with the principal point chain closest to length in pixels be more than 8
Fulcrum chain；
It is separated by most nearby in principal point chain and fulcrum chain, takes a neighborhood as ROI, takes the part of principal point chain in ROI respectively
Fitting a straight line is carried out with the part of fulcrum chain；
Judge the position relationship of principal point chain and fulcrum chain：
If the local fit straight line of principal point chain intersects in itself with fulcrum chain, the local fit along the principal point chain is straight
Line extends principal point chain to be connected with fulcrum chain, and the bearing of trend of the blood vessel characteristic point on connecting line is the angle of connection line slope；
If the local fit straight line of fulcrum chain intersects in itself with principal point chain, the local fit along the fulcrum chain is straight
Line extends fulcrum chain to be connected with principal point chain, and the bearing of trend of the blood vessel characteristic point on connecting line is the angle of connection line slope.
On the other hand, a kind of vein distribution display device, described device include：
Image acquisition units, for gathering the infrared image and visible images at position to be detected；
Intermediate image acquiring unit, for being pre-processed to the infrared image, to remove noise and strengthen contrast,
Obtain intermediate image；
Characteristic point image acquisition unit, for asking for the curvature maximum of the intermediate image, and noise spot is filtered out, with
To the feature dot image for including all blood vessel characteristic points, wherein, the result images are included corresponding to each blood vessel characteristic point
Blood vessel bearing of trend；
Vessel branch image acquisition unit, for based on the blood vessel bearing of trend, to the blood in the feature dot image
Pipe characteristic point is attached processing, obtains vessel branch image；
Center line image acquisition unit, for the vessel branch in the vessel branch image to be connected into vein network,
To form the center line of human vas, and the burr on the center line is removed, the center line image after being handled；
Blood vessel color filling unit, for determining the target blood region in the center line image, and described visible
It is corresponding with the target blood region in light image to be partially filled with blood vessel color.
Beneficial effects of the present invention comprise at least：
The present invention on the premise of accurate display mainline distribution is realized, protect completely by the epidermis for realizing non-vein region
The purpose of skin primary colors has been stayed, the missing image shown by traditional vein display system has been solved or dermatogen can not be recovered completely
The problem of color.Doctor is set more true, effectively to observe that vein is distributed in venipuncture or other diagnosis and treatment processes.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, make required in being described below to embodiment
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for
For those of ordinary skill in the art, on the premise of not paying creative work, other can also be obtained according to these accompanying drawings
Accompanying drawing.
Fig. 1 is embodiment of the present invention medium sized vein distribution display method flow chart；
Fig. 2 is the functional block diagram of embodiment of the present invention medium sized vein distribution display device.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is described in further detail.
Referring to Fig. 1, the application provides a kind of vein distribution display method, the described method comprises the following steps：
S11：Gather the infrared image and visible images at position to be detected；
S12：The infrared image is pre-processed, to remove noise and strengthen contrast, obtains intermediate image；
S13：The curvature maximum of the intermediate image is asked for, and filters out noise spot, to obtain including all blood vessel features
The feature dot image of point, wherein, the result images include blood vessel bearing of trend corresponding to each blood vessel characteristic point；
S14：Based on the blood vessel bearing of trend, processing is attached to the blood vessel characteristic point in the feature dot image,
Obtain vessel branch image；
S15：Vessel branch in the vessel branch image is connected into vein network, to form the center of human vas
Line, and the burr on the center line is removed, the center line image after being handled；
S16：Determine the target blood region in the center line image, and in the visible images with the target
Blood vessel color is partially filled with corresponding to angiosomes.
A kind of specific implementation step of vein distribution display method provided by the invention is as follows：
Step S1：Gather the infrared image and visible images at position to be detected；
Step S2：Image registration is carried out according to the infrared image obtained in Same Scene and coloured image；
On four angles of target area respectively one additional character of mark (such as cross angle point) as registration feature
Image.
By infrared and visible images discoloration, it is converted into gray level image and takes four special symbols of two width gray level images respectively
Number, with the coordinate of the characteristic point of Harris operators (or operator of other similar effects) four additional characters of extraction.
The infrared and registration region of visible ray two images can be calculated in given imaging plane with this two groups of coordinates
Mutual mapping relations matrix.
Using the infrared registration region image Jing Guo Edge Gradient Feature as reference picture, with by Edge Gradient Feature
Visible ray registration region image enters line translation to visible images with mapping relations matrix, makes two width figures as image subject to registration
Image position is corresponding.
Step S3：Infrared image is pre-processed, noise is filtered out and improves contrast, exports intermediate image；
The Otsu threshold of infrared image is calculated, the approximate region of human body skin in image is extracted using Otsu threshold；
The definition of Otsu threshold is：The histogram of image to be split is provided, a threshold value t is calculated in traversal so that should
The inter-class variance of threshold value both sides is maximum, i.e.,：
F (t)=pixnum (I | I ＞ t) (ave (I | I ＞ t)-ave (I))2+ pixnum (I | I ＜ t) (ave (I | I
＜ t)-ave (I))2T value when obtaining maximum；
Pixnum is pixel count of the gray value in the range of this in formula, and ave is pixel average；
After obtaining the approximate region of human body skin, contrast enhancing individually is carried out to this area-of-interest I_ROI：
Then enhanced image noml is filtered using gaussian filtering：
F (x, y)=nom1 (x, y) * GaussianKernel (11,2.5)；
GaussianKernel is Gaussian kernel in formula, and the operator length of side should be less than the width (value herein of image medium vessels
11), σ=2.5；
Contrast decreases after filtering, therefore also to carry out a contrast enhancing operation：
F (x, y) be source images value, mean be source images average value, V be source images variance, MsetTo be default defeated
Go out image average, VsetFor the variance of default output image；
As F (x, y) ＜ mean,
As F (x, y) ＞ mean,
The intermediate image of denoising and contrast enhancing after being pre-processed.
Step S4：The curvature maximum of intermediate image is asked for, and filters out noise spot, a width is obtained and includes all blood vessel centers
The image of line feature point, and each characteristic point correspond to the bearing of trend of blood vessel；
A line is chosen on image P (x, y) first, calculates its curvature
All local maximums of this line curvature are found out again.Its decision condition is：
1st, curvature be necessary on the occasion of；
2nd, in often going, Local modulus maxima is bordered by relation being a little present with it
And
3rd, Local modulus maxima both sides monotonicity is different
ρ(xmax) ＜ ρ (xmax- 1) ＜ ρ (xmax-2)
And ρ (xmax) ＜ ρ (xmax+ 1) ＜ ρ (xmax+2)
4th, the point for meeting above three condition is referred to as curvature peak point, it is believed that (curvature is more than 0 to the full duration of peak value both sides
Scope) be blood vessel width；In order to exclude the interference of skin lines and hair, if the full duration of some curvature peak point is less than
8, then it is assumed that this peak point is noise spot；
Every a line, each row and the oblique line along 45 °, 135 ° directions of image are processed as, obtains four groups of local maximums
Point.
Will per a line (0 °), each row (90 °) and four groups of maximum point sinks along 45 °, 135 ° directions are neat, make in a figure
On Pp；
Write down the bearing of trend of each characteristic point in addition simultaneously, bearing of trend mutually hangs down with extracting the direction of this feature point
Directly.
To Pp filtering and noise reductions：Scatterplot (not having the point of other characteristic points in the range of 3*3) is removed first, is next removed discrete
In the circumvascular point of big section (less than the point of 7 characteristic points in the range of 19*19 around the point), a scatterplot is finally removed again
(there is no the point of other characteristic points in the range of 5*5)；
Previous step has some characteristic points also to be left out after terminating, and therefore, is searched in the range of remaining characteristic point 7*7
The characteristic point erased, then add back, obtain final feature dot image P；
Step S5：Characteristic point inside connection features dot image P, fill a vacancy, form vessel branch.
The step first stage, it is the vacancy that connection only has a pixel；
(1) all vessel segments in image are extracted：The connected characteristic point of all composition 8- syntoples can be regarded as one group.Extract institute
Some group, chain (equivalent to n point coordinates array is obtained, n is point chain number) is hereinafter referred to as put, a chain is numbered.Point chain
As vessel centerline；
(2) all point chains pair for being only separated by a pixel are counted；
(3) these chains pair are most nearby being connected, the bearing of trend of tie point is the characteristic point bearing of trend angle of both sides
Average value；
The step second stage, be the larger vacancy on the same branch vessel center line of connection, can also connect be separated by it is nearer
Two branch vessels；
(1) having a chain and numbering in image is extracted first, and the end points for finding a chain is marked；
(2) longer point chain (length in pixels is more than 8 point chain) is therefrom found out, writes down its numbering；
(3) (threshold value is set to no more than the length threshold of above long point chain in every long chain periphery certain distance of searching
8) other chains, the numbering of these chains is attached to after the long point chain numbering；
Travel through all long point chains, every point chain by the following situation analysis chain and around it in 8 pixel distances：
Situation 1：If two point chains are all long point chains, compare the angle of two point chain fitting a straight lines；
Fitting a straight line is carried out to two point chains respectively first：
Fitting is divided into two steps：The first step is that a chain is screened using consistent (RANSAC) method of random sampling, is obtained
Point chain after denoising；
Second step is to carry out fitting a straight line to the point chain after denoising using the method for singular value decomposition.Chain s set up an office by n point
Form, i-th point of coordinate is (xi, yi), then it can obtain the point range of homogeneous partial differential：
IfThen obtain transition matrix
Then first s can be normalized：
Sn=Ts
Singular value decomposition then is carried out to sn
snT=U. σ .VT
Now V the third line is the parameter of linear equation.
If the angle of two fitting a straight lines is more than 45 °, the intersection point of two fitting a straight lines is made, two point chains connect respectively
It is connected on this intersection point；
Otherwise, the closest approach of two point chains is directly taken to be connected, the bearing of trend of the characteristic point on connecting line is oblique for connecting line
The angle of rate.
Situation 2：If one is long and the other is short for two point chains, judge short chain whether in fitting a straight line (the fitting side of long point chain
Method is same as above) direction on, judgment basis is：
Short chain is less than 30 degree to the line of long point chain and the angle of long point chain fitting a straight line, and short chain is in the long point end of the chain
Point then takes the closest approach of 2 chains to be connected, the bearing of trend of the characteristic point on connecting line is nearby rather than near middle part
Connect the angle of line slope；Otherwise it is not attached to.
From (1) step of second stage to this step, if chain sum in image midpoint has been reduced, again from second stage
(1) step the step of starting to perform larger vacancy connection, after performing a flow, image midpoint chain sum no longer subtracts
It is few.
So far, it is believed that all vessel branches have all had a style of one's own, and obtain connecting the vessel branch image after breakpoint
P1。
Step S6：Vessel branch in P1 is connected into vein network, forms the center line of human vas, in then removing
Burr branch on heart line, obtain the vein network center line image P2 after connection processing：
Take point chain most long in now image (hereinafter referred to as principal point chain), look for its closest to a long point chain (hereinafter referred to as
Fulcrum chain)；
It is separated by most nearby in principal point chain and fulcrum chain, takes a neighborhood as ROI, takes the part of principal point chain in ROI respectively
Fitting a straight line is carried out with the part of fulcrum chain；
Judge the position relationship of principal point chain and fulcrum chain：
If the local fit straight line of principal point chain intersects in itself with fulcrum chain, extend principal point chain and fulcrum along this line
Chain is connected, and the bearing of trend of the characteristic point on connecting line is the angle of connection line slope；
, whereas if the local fit straight line of fulcrum chain intersects in itself with principal point chain, then extend fulcrum chain along this line
It is connected with principal point chain, the bearing of trend of the characteristic point on connecting line is the angle of connection line slope.
From " take point chain most long in now image (hereinafter referred to as principal point chain), look for its closest to a long point chain " the step of
To this step, if chain sum in image midpoint has been reduced, again from " taking point chain most long in now image (hereinafter referred to as
Principal point chain), look for its closest to a long point chain " the step of start the step of execution connects into vein network, until performing once
After flow, image midpoint chain sum is no longer reduced.
Take point chain most long in now image, as the back of the hand mainline blood vessel network center line；
Some shorter burr bifurcateds (length threshold 8) on mainline blood vessel network center line are removed, obtain P2.
Step S7：The width of blood vessel is extracted, and calculates mean breadth；
Each characteristic point on P2 has bearing of trend, and the direction vertical with bearing of trend is blood vessel width.
Searched for along width, obtain the spike width that this feature point radius of curvature is more than 0, the width as the blood vessel
Degree；
All characteristic points have corresponding blood vessel width, and it is averaged, and obtain blood vessel mean breadth
R in formulaiFor the width of ith feature point, n is characterized a sum.
Step S8：Using expansion algorithm, vessel region of interest V is obtained：
With blood vessel widthFor diameter, circular expander operator is built；
Using the Expanded Operators of previous step, expansive working is performed to P2, obtains vessel region of interest V；
Step S9：Vessel region of interest is painted on visible images Vis, reaches the effect of enhancing display.
To angiosomes, using the color (Ra close to blue veins：51, Ga：87, Ba：65) pixel for combining colouring region is strong
Degree is adjusted colouring；
Depending on the display color of the pixel of angiosomes is by following formula：
X is the pixel intensity value of infrared image in formula, and a is average value of the infrared image in vessel region of interest V,
Var is variance of the infrared image in vessel region of interest V, and Xa is color (X references R, G, B tri- of default three passages
Individual passage), var ' is the variance (value 500) after default change, and Xs is the color of three passages finally shown.
To visible images Vis, the part in vessel region of interest V uses Xs three passages colouring, its remaining part instead
Divide constant：
Vis ((x, y) | (x, y) ∈ V)=Xs
So, finally display image can be strengthened to obtain vein.
Referring to Fig. 2, the application also provides a kind of vein distribution display device, described device includes：
Image acquisition units 100, for gathering the infrared image and visible images at position to be detected；
Intermediate image acquiring unit 200, for being pre-processed to the infrared image, to remove noise and strengthen contrast
Degree, obtains intermediate image；
Characteristic point image acquisition unit 300, for asking for the curvature maximum of the intermediate image, and noise spot is filtered out,
To obtain including the feature dot image of all blood vessel characteristic points, wherein, the result images include each blood vessel characteristic point pair
The blood vessel bearing of trend answered；
Vessel branch image acquisition unit 400, for based on the blood vessel bearing of trend, in the feature dot image
Blood vessel characteristic point is attached processing, obtains vessel branch image；
Center line image acquisition unit 500, for the vessel branch in the vessel branch image to be connected into rete venosum
Network, to form the center line of human vas, and the burr on the center line is removed, the center line image after being handled；
Blood vessel color filling unit 600, for determining the target blood region in the center line image, and it is described can
See and corresponding with the target blood region in light image be partially filled with blood vessel color.
The specific steps that unit is realized in the vein distribution display device that the application provides, are distributed with above-mentioned vein
Description in display methods is consistent, just repeats no more here.
Beneficial effects of the present invention comprise at least：
The present invention on the premise of accurate display mainline distribution is realized, protect completely by the epidermis for realizing non-vein region
The purpose of skin primary colors has been stayed, the missing image shown by traditional vein display system has been solved or dermatogen can not be recovered completely
The problem of color.Doctor is set more true, effectively to observe that vein is distributed in venipuncture or other diagnosis and treatment processes.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
Realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Based on such understanding, on
The part that technical scheme substantially in other words contributes to prior art is stated to embody in the form of software product, should
Computer software product can store in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including some fingers
Make to cause a computer equipment (can be personal computer, server, or network equipment etc.) to perform each implementation
Method described in some parts of example or embodiment.
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc., it should be included in the scope of the protection.
Claims (10)
1. a kind of vein distribution display method, it is characterised in that methods described includes：
Gather the infrared image and visible images at position to be detected；
The infrared image is pre-processed, to remove noise and strengthen contrast, obtains intermediate image；
The curvature maximum of the intermediate image is asked for, and filters out noise spot, to obtain including the feature of all blood vessel characteristic points
Dot image, wherein, the result images include blood vessel bearing of trend corresponding to each blood vessel characteristic point；
Based on the blood vessel bearing of trend, processing is attached to the blood vessel characteristic point in the feature dot image, obtains blood vessel
Branching diagram picture；
Vessel branch in the vessel branch image is connected into vein network, to form the center line of human vas, and gone
Except the burr on the center line, the center line image after being handled；
Determine the target blood region in the center line image, and in the visible images with the target blood region
It is corresponding to be partially filled with blood vessel color.
2. according to the method for claim 1, it is characterised in that gathering the infrared image and visible ray figure at position to be detected
As after, methods described also includes：
The infrared image and the visible images are subjected to registration, so that the position of two images is corresponding.
3. according to the method for claim 1, it is characterised in that the blood vessel characteristic point in the feature dot image is connected
Connecing processing includes：
Two blood vessel characteristic points of the vacancy at a distance of only one pixel are connected first, then according to blood corresponding to blood vessel characteristic point
The vessel centerline of pipe bearing of trend connection interruption, is linked to be vessel branch by the blood vessel characteristic point scattered.
4. according to the method for claim 1, it is characterised in that determine the target blood region bag in the center line image
Include：
The two-sided search curvature of the bearing of trend of blood vessel characteristic point is more than zero scope on heart line in the blood vessel, and search is obtained
Width of the range wide as blood vessel at this, and calculate the mean breadth of blood vessel；
Expansive working is carried out to the center line image with the mean breadth, obtains target blood region.
5. according to the method for claim 1, it is characterised in that the intermediate image determines in the following manner：
The Otsu threshold of the infrared image is calculated, and the area of human body skin in the infrared image is extracted using Otsu threshold
Domain；
Contrast enhancing is carried out to the region of the human body skin according to the following equation：
<mrow>
<mi>n</mi>
<mi>o</mi>
<mi>m</mi>
<mn>1</mn>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mn>255</mn>
<mo>*</mo>
<mfrac>
<mrow>
<mo>(</mo>
<mi>I</mi>
<mo>_</mo>
<mi>R</mi>
<mi>O</mi>
<mi>I</mi>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
<mo>-</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
<mo>(</mo>
<mi>I</mi>
<mo>_</mo>
<mi>R</mi>
<mi>O</mi>
<mi>I</mi>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mo>(</mo>
<mi>I</mi>
<mo>_</mo>
<mi>R</mi>
<mi>O</mi>
<mi>I</mi>
<mo>)</mo>
<mo>-</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
<mo>(</mo>
<mi>I</mi>
<mo>_</mo>
<mi>R</mi>
<mi>O</mi>
<mi>I</mi>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mfrac>
</mrow>
Wherein, nom1 (x, y) represents pixel value corresponding to the region of the enhanced human body skin of contrast, and I_ROI (x, y) is represented
Pixel value corresponding to the region of human body skin before contrast enhancing, min expressions take minimum operation, and max represents to take maximum
It is worth computing, (x, y) represents the coordinate of pixel；
Pixel value corresponding to the region of the enhanced human body skin of the contrast is filtered according to the following equation：
F (x, y)=nom1 (x, y) * GaussianKernel (11,2.5)；
Wherein, GaussianKernel is Gaussian kernel, and F (x, y) is filtered pixel value；
Carry out contrast enhancing operation again to the filtered pixel value according to the following equation：
As F (x, y) ＜ mean,
<mrow>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>s</mi>
<mi>e</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>-</mo>
<msqrt>
<mfrac>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<mi>F</mi>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
<mo>-</mo>
<mi>m</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>&times;</mo>
<msub>
<mi>V</mi>
<mrow>
<mi>s</mi>
<mi>e</mi>
<mi>t</mi>
</mrow>
</msub>
</mrow>
<mi>V</mi>
</mfrac>
</msqrt>
</mrow>
As F (x, y) ＞ mean,
<mrow>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>s</mi>
<mi>e</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>+</mo>
<msqrt>
<mfrac>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<mi>F</mi>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
<mo>-</mo>
<mi>m</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>&times;</mo>
<msub>
<mi>V</mi>
<mrow>
<mi>s</mi>
<mi>e</mi>
<mi>t</mi>
</mrow>
</msub>
</mrow>
<mi>V</mi>
</mfrac>
</msqrt>
</mrow>
Wherein, mean is pixel average, and V is the variance of pixel value, MsetFor default output image average, VsetTo be default
The variance of output image, I1(x, y) represents to carry out the enhanced pixel value of contrast again；
Using the image for carrying out the enhanced pixel value composition of contrast again as the intermediate image.
6. according to the method for claim 5, it is characterised in that the Otsu threshold determines in the following manner：
The histogram of image to be split is provided, a threshold value t is calculated in traversal so that the inter-class variance of threshold value t both sides is most
Greatly, i.e.,：
Pixnum (I | I ＞ t) (ave (I | I ＞ t)-ave (I))2+ pixnum (I | I ＜ t) (ave (I | I ＜ t)-ave (I)
)2When obtaining maximum, the value of the threshold value t is determined；
Wherein, pixnum is the pixel count of gray value within the specified range, and ave is pixel average, and I represents the infrared image
Pixel value.
7. according to the method for claim 1, it is characterised in that the blood vessel characteristic point in the feature dot image is connected
Connecing processing includes：
All vessel segments in the feature dot image are extracted, the connected blood vessels characteristic point of all composition 8- syntoples can be regarded as a point
Chain；
All point chains are extracted, a chain is numbered；Wherein, described chain characterizes vessel centerline；
Count all point chains pair for being only separated by a pixel；
In the point chain pair that most nearby connection statistics go out, the bearing of trend of tie point is the blood vessel characteristic point bearing of trend angle of both sides
Average value.
8. according to the method for claim 7, it is characterised in that methods described also includes：
Having a chain and numbering in the feature dot image is extracted, the end points for finding a chain is marked；
Long point chain of the length in pixels more than 8 is found out from described chain, and writes down the long point chain numbering；
Other chains in every long chain periphery distance to a declared goal are found, and the numbering of the point chain searched out is attached to the long point chain
After numbering.
9. according to the method for claim 1, it is characterised in that the center line image after determination processing in the following manner：
Take principal point chain most long in vessel branch image, and find with the principal point chain closest to length in pixels be more than 8 branch
Point chain；
It is separated by most nearby in principal point chain and fulcrum chain, takes a neighborhood to take the part of principal point chain and branch in ROI respectively as ROI
The part of point chain carries out fitting a straight line；
Judge the position relationship of principal point chain and fulcrum chain：
If the local fit straight line of principal point chain intersects in itself with fulcrum chain, the local fit straight line along the principal point chain prolongs
For long principal point chain to be connected with fulcrum chain, the bearing of trend of the blood vessel characteristic point on connecting line is the angle of connection line slope；
If the local fit straight line of fulcrum chain intersects in itself with principal point chain, the local fit straight line along the fulcrum chain prolongs
For long fulcrum chain to be connected with principal point chain, the bearing of trend of the blood vessel characteristic point on connecting line is the angle of connection line slope.
10. a kind of vein is distributed display device, it is characterised in that described device includes：
Image acquisition units, for gathering the infrared image and visible images at position to be detected；
Intermediate image acquiring unit, for being pre-processed to the infrared image, to remove noise and strengthen contrast, obtain
Intermediate image；
Characteristic point image acquisition unit, for asking for the curvature maximum of the intermediate image, and noise spot is filtered out, to be wrapped
Feature dot image containing all blood vessel characteristic points, wherein, the result images include blood vessel corresponding to each blood vessel characteristic point
Bearing of trend；
Vessel branch image acquisition unit, it is special to the blood vessel in the feature dot image for based on the blood vessel bearing of trend
Sign point carries out junction reason, obtains vessel branch image；
Center line image acquisition unit, for the vessel branch in the vessel branch image to be connected into vein network, with structure
The center line of adult body blood vessel, and the burr on the center line is removed, the center line image after being handled；
Blood vessel color filling unit, for determining the target blood region in the center line image, and in the visible ray figure
It is corresponding with the target blood region as in be partially filled with blood vessel color.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710804444.4A CN107749049B (en) | 2017-09-07 | 2017-09-07 | Vein distribution display method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710804444.4A CN107749049B (en) | 2017-09-07 | 2017-09-07 | Vein distribution display method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107749049A true CN107749049A (en) | 2018-03-02 |
| CN107749049B CN107749049B (en) | 2021-09-28 |
Family
ID=61254876
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710804444.4A Active CN107749049B (en) | 2017-09-07 | 2017-09-07 | Vein distribution display method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107749049B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145748A (en) * | 2018-07-23 | 2019-01-04 | 南京航空航天大学 | The intravenous line extraction algorithm of visible light vein visualising image |
| CN109886973A (en) * | 2019-01-25 | 2019-06-14 | 杭州晟视科技有限公司 | A kind of vessel extraction method, apparatus and computer readable storage medium |
| CN110197495A (en) * | 2019-05-30 | 2019-09-03 | 数坤（北京）网络科技有限公司 | The method of adjustment and device of vessel extraction |
| CN112529800A (en) * | 2020-12-07 | 2021-03-19 | 同济大学 | Near-infrared vein image processing method for filtering hair noise |
| CN112907548A (en) * | 2021-02-26 | 2021-06-04 | 依未科技(北京)有限公司 | Image evaluation method and device, computer-readable storage medium and electronic device |
| CN113011333A (en) * | 2021-03-19 | 2021-06-22 | 哈尔滨工业大学 | System and method for obtaining optimal venipuncture point and direction based on near-infrared image |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103417196A (en) * | 2013-08-23 | 2013-12-04 | 中山大学 | Venous visualizer and visualizing method |
| CN104688184A (en) * | 2014-12-05 | 2015-06-10 | 南京航空航天大学 | Vein imaging method for visible-light skin images |
| CN105125176A (en) * | 2015-09-23 | 2015-12-09 | 博联众科（武汉）科技有限公司 | Device and method for vein navigation |
-
2017
- 2017-09-07 CN CN201710804444.4A patent/CN107749049B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103417196A (en) * | 2013-08-23 | 2013-12-04 | 中山大学 | Venous visualizer and visualizing method |
| CN104688184A (en) * | 2014-12-05 | 2015-06-10 | 南京航空航天大学 | Vein imaging method for visible-light skin images |
| CN105125176A (en) * | 2015-09-23 | 2015-12-09 | 博联众科（武汉）科技有限公司 | Device and method for vein navigation |
Non-Patent Citations (3)
| Title |
|---|
| FOTIOS TAGKALAKIS ET AL.: "A low cost Finger Vein Authentication System,using Maximum Curvature Points", 《UNIVERSITY OF WEST BOHEMIA》 * |
| XIANJING MENG ET AL.: "Finger Vein Recognition Based on Local Directional Code", 《SENSORS 2012》 * |
| 孙晓琳: "指静脉图像模式提取算法研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145748A (en) * | 2018-07-23 | 2019-01-04 | 南京航空航天大学 | The intravenous line extraction algorithm of visible light vein visualising image |
| CN109145748B (en) * | 2018-07-23 | 2021-10-26 | 南京航空航天大学 | Vein line extraction algorithm of visible light vein imaging image |
| CN109886973A (en) * | 2019-01-25 | 2019-06-14 | 杭州晟视科技有限公司 | A kind of vessel extraction method, apparatus and computer readable storage medium |
| CN109886973B (en) * | 2019-01-25 | 2021-01-08 | 杭州晟视科技有限公司 | Blood vessel extraction method and device and computer readable storage medium |
| CN110197495A (en) * | 2019-05-30 | 2019-09-03 | 数坤（北京）网络科技有限公司 | The method of adjustment and device of vessel extraction |
| CN110197495B (en) * | 2019-05-30 | 2021-03-09 | 数坤（北京）网络科技有限公司 | Adjusting method and device for blood vessel extraction |
| CN112529800A (en) * | 2020-12-07 | 2021-03-19 | 同济大学 | Near-infrared vein image processing method for filtering hair noise |
| CN112529800B (en) * | 2020-12-07 | 2022-08-23 | 同济大学 | Near-infrared vein image processing method for filtering hair noise |
| CN112907548A (en) * | 2021-02-26 | 2021-06-04 | 依未科技(北京)有限公司 | Image evaluation method and device, computer-readable storage medium and electronic device |
| CN113011333A (en) * | 2021-03-19 | 2021-06-22 | 哈尔滨工业大学 | System and method for obtaining optimal venipuncture point and direction based on near-infrared image |
| CN113011333B (en) * | 2021-03-19 | 2022-11-18 | 哈尔滨工业大学 | System and method for obtaining optimal venipuncture point and direction based on near-infrared image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107749049B (en) | 2021-09-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107749049A (en) | A kind of vein distribution display method and device | |
| CN105844655B (en) | A kind of laser rays stripe edge extracting method | |
| CN107564048B (en) | Feature registration method based on bifurcation point | |
| CN102521859B (en) | Reality augmenting method and device on basis of artificial targets | |
| CN108510502A (en) | Melanoma picture tissue segmentation methods based on deep neural network and system | |
| CN107066957A (en) | Iris locating method and device in visible ray eyes image | |
| WO2020038312A1 (en) | Multi-channel tongue body edge detection device and method, and storage medium | |
| US20180220893A1 (en) | Region of interest tracing apparatus | |
| CN108154147A (en) | The region of interest area detecting method of view-based access control model attention model | |
| CN109325421A (en) | A kind of eyelashes minimizing technology and system based on edge detection | |
| CN106203451A (en) | A kind of image area characteristics extracts and the method for characteristic matching | |
| CN111340773A (en) | Retinal image blood vessel segmentation method | |
| CN106408533A (en) | Card image extraction method and card image extraction system | |
| CN105631456A (en) | Particle swarm optimization ITTI model-based white cell region extraction method | |
| CN108734054A (en) | Unobstructed citrusfruit image-recognizing method | |
| CN107239761A (en) | Fruit tree branch pulling effect evaluation method based on skeleton Corner Detection | |
| CN107705303A (en) | The dividing method of blood vessel on a kind of medical image | |
| Goldbaum et al. | Automated registration of digital ocular fundus images for comparison of lesions | |
| CN104008365A (en) | Method for detecting sparse degree of fruit tree leaves based on image processing technology | |
| CN104616260A (en) | Vein image enhancement method and device | |
| CN110929681A (en) | Wrinkle detection method | |
| Lau et al. | Exploring edge-based segmentation towards automated skin lesion diagnosis | |
| CN104537350A (en) | License plate positioning method based on multi-dimensional edge feature | |
| Bian et al. | Research of face detection system based on skin-tone feature | |
| Agnalt et al. | Vein detection using vein transillumination and contrast differentiation for practitioner aid |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |