CN109377549A - A kind of real-time processing of OCT finger tip data and three-dimensional visualization method - Google Patents
A kind of real-time processing of OCT finger tip data and three-dimensional visualization method Download PDFInfo
- Publication number
- CN109377549A CN109377549A CN201811145157.8A CN201811145157A CN109377549A CN 109377549 A CN109377549 A CN 109377549A CN 201811145157 A CN201811145157 A CN 201811145157A CN 109377549 A CN109377549 A CN 109377549A
- Authority
- CN
- China
- Prior art keywords
- data
- finger tip
- scan
- oct
- gpu
- 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
- 238000012545 processing Methods 0.000 title claims abstract description 32
- 238000007794 visualization technique Methods 0.000 title claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 30
- 238000003384 imaging method Methods 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 15
- 238000012800 visualization Methods 0.000 claims abstract description 12
- 210000000106 sweat gland Anatomy 0.000 claims abstract description 9
- 230000006870 function Effects 0.000 claims description 23
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 claims description 14
- 210000003491 skin Anatomy 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 239000000835 fiber Substances 0.000 claims description 7
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 4
- 238000012952 Resampling Methods 0.000 claims description 3
- 230000001133 acceleration Effects 0.000 claims description 3
- 210000002615 epidermis Anatomy 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 239000011800 void material Substances 0.000 claims description 3
- 230000008878 coupling Effects 0.000 claims description 2
- 238000010168 coupling process Methods 0.000 claims description 2
- 238000005859 coupling reaction Methods 0.000 claims description 2
- 206010033675 panniculitis Diseases 0.000 abstract description 6
- 210000004304 subcutaneous tissue Anatomy 0.000 abstract description 6
- 210000001519 tissue Anatomy 0.000 abstract description 6
- 238000012014 optical coherence tomography Methods 0.000 description 26
- 238000009877 rendering Methods 0.000 description 7
- 238000005070 sampling Methods 0.000 description 5
- 238000007920 subcutaneous administration Methods 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 230000001427 coherent effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000005266 casting Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000005610 quantum mechanics Effects 0.000 description 2
- 238000003786 synthesis reaction Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000037311 normal skin Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000010008 shearing Methods 0.000 description 1
- 238000010408 sweeping Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/08—Indexing scheme for image data processing or generation, in general involving all processing steps from image acquisition to 3D model generation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A kind of real-time processing of OCT finger tip data and three-dimensional visualization method include the following steps: 1) to obtain three-dimensional original bin data using frequency domain OCT system scanning finger sample；2) according to interference imaging principle, parsing imaging is carried out to initial data, treatment process is accelerated by the high-speed parallel operational capability of GPU, the real-time processing of data is realized and data is stored in three-dimensional array by treated；3) using improved light projecting algorithm, to treated, finger tip data carry out three-dimensional visualization, show fingerprint in the Epidermal Fingerprint, sweat gland and skin corium of finger tip.The present invention facilitates more accurate fingerprint recognition；The treatment process of OCT finger tip data is accelerated by the high-speed parallel operational capability of GPU, can be realized real-time processing；Three-dimensional visualization is carried out to treated finger tip data, can more intuitively and accurately show the spatial relationship of the spatial position of subcutaneous tissue of finger, size, geometry and it and surrounding tissue.
Description
Technical field
Real-time processing and three-dimensional visualization method the present invention relates to a kind of OCT finger tip data, and in particular to GPU is counted parallel
Calculation and Volume Rendering Techniques.
Background technique
Optical Coherence Tomography Imaging Technology (Optical Coherence Tomograph, OCT) is a kind of novel optics
Imaging technique, it utilizes the basic principle of weak coherent light interferometer, and detection biological tissue's different depth level is to incident weak relevant
The backscatter signals of light can obtain the two dimension or three-dimensional structure image of biological tissue.It uses infrared light scanning, penetrates not
Deep, imaging depth only has 2-3mm in general biological tissue, but 1000 times of its identification capability ratio CT high, and resolution ratio can
Reach 1 μm~15 μm, be applied to the high-resolution imaging of eye, skin, tooth etc., is that doctor carries out clinical diagnosis and treatment
Strong supplementary means.In recent years, OCT is gradually applied on finger biometric collection apparatus, and traditional finger characteristic acquisition is main logical
The mode for crossing pressing directly acquires two-dimensional fingerprint image, then according to the global characteristics of fingerprint image and minutia come to people
Identity identified that and OCT is deep into subcutaneous tissue from air incidence to finger epidermis then by weak coherent light again, in addition to
Epidermal Fingerprint is acquired, subcutaneous sweat gland, these new biological informations of fingerprint in skin corium are also acquired.
GPU (Graphics Processing Unit), i.e. graphics processor, internal a large amount of transistor are used to do number
According to operation, there is great advantage in the parallel computation of mass data.CUDA(Computer Unified Device
It Architecture is) NVIDIA in the programming API and framework that 2006 are that the multiple programming on GPU provides, in CUDA model
In, CPU is responsible for the distribution, data transmission and the starting of kernel of memory and video memory as host (Host), GPU as association at
Device or equipment (Device) are managed, the calculating of a large amount of Method on Dense Type of Data Using of parallel section is responsible for.
Volume Rendering Techniques (Volume Rendering Technique) are directly to be generated by discrete 3 d data field pair
Answer a kind of rendering technique of two dimensional image.Different from iso-surface patch, it is without generating intermediate geometric graphic element.Direct volume drawing
The representative algorithm of (Direct Volume rendering) mainly includes light quantum mechanics (Ray-Casting), maximum intensity throwing
Shadow algorithm (Maximum Intensity Projection), snow throwing ball (Splatting) and shearing Qu Bianfa (Shear-
Warp) etc..Wherein light quantum mechanics (Ray-Casting) are the classical rendering algorithms of image space, it is from the every of projection plane
A point issues throw light, passes through 3 d data field, calculates light intensity and drafting pattern picture after decaying by equations of light ray,
Its rendering quality highest, but speed is slower, but as development this problem of graphic hardware is gradually resolved.
With stepping up for OCT sampling rate, data collection capacity is huge, and data handling procedure includes many times FFT
(Fast Fourier Transform) transformation, the algorithm computation complexity is high, and entire data handling procedure is time-consuming serious.
Summary of the invention
It is an object of the invention to explore application problem of the OCT technology on physical characteristics collecting, by GPU high speed simultaneously
Row operational capability accelerates the treatment process of OCT finger tip data, realizes the real-time processing of finger tip data, and with improved
Light projecting algorithm shows the three-dimensional structure of subcutaneous tissue of finger, and the biology so as to preferably reflect subcutaneous tissue of finger is special
Reference breath.
The purpose of the present invention is achieved through the following technical solutions:
A kind of real-time processing of OCT finger tip data and three-dimensional visualization method, include the following steps:
1) three-dimensional original bin data is obtained using frequency domain OCT system scanning finger sample；
2) according to interference imaging principle, parsing imaging is carried out to initial data, by the high-speed parallel operational capability pair of GPU
Treatment process is accelerated, and realizes the real-time processing of data and data are stored in three-dimensional array by treated；
3) using improved light projecting algorithm, to treated, finger tip data carry out three-dimensional visualization, show the table of finger tip
Fingerprint in skin fingerprint, sweat gland and skin corium.
Further, in the step 1), the process of the frequency domain OCT system acquisition finger tip data: wideband light source SLD hair
Low-coherent light (central wavelength lambda 848nm, bandwidth Delta lambda 46nm) out is divided into two-beam through four end fiber couplers, with 50/
50 ratio respectively enters reference arm and sample arm；It is incident in parallel into the light in reference arm after collimation lens collimates
Onto total reflection mirror, and reflected in parallel as reference light；It is focused on sample into the light in sample arm, by sweeping
It retouches galvanometer system to be scanned finger sample, is reflected as signal light by the scattering particles of different depth in finger sample；From
The reference light and signal light returned in reference arm and sample arm converges in fiber coupler, and superposition and interference, interference letter occurs
Number from the coupler other end outgoing enter spectrometer parsed, acquired by line array CCD, computer be sent by data collecting card
Middle to carry out a series of data processing, then by treated, data carry out 2D or 3D imaging.
The step of acquisition finger tip data, is as follows:
1.1) an axial scan A-SCAN is carried out to a point on finger tip, it includes depth information that CCD, which collects one,
Line spectrum；
1.2) transversal scanning B-SCAN is carried out to finger tip, obtains the depth information spectrum of a plurality of A-SCAN composition；
1.3) the individual scan data of transversal scanning available one is successively repeatedly carried out backward.
The axial resolution of frequency domain OCT system used in the present invention is 7 μm, and lateral resolution is 11 μm, the pixel of CCD
It pair is 2048, then the number of pixels that an A-SCAN is obtained is 2048, and B-SCAN includes 500 A-SCAN signals, i.e.,
Finger sample a direction has carried out the acquisition of 500 depth datas, and sweep speed is 36klines/s, each B-SCAN size
For 500lines × 2048pixels/line × 2bytes/pixel, amount to 1.9MB data, an individual data items are by many B-
SCAN composition, data volume are very big.
In the step 2), data handling procedure includes being averaging to a plurality of A-SCAN, and every A-SCAN is counted respectively
According to interception, interpolation, subtracts direct current, FFT transform, modulus and take these operations of logarithm；Other than being averaging, other operations are all one
What mono- A-SCAN of A-SCAN was individually carried out, a large amount of single data elements can be calculated with complete independently, be suitble on GPU parallel
Processing, therefore the acceleration to data handling procedure can be realized by the high-speed parallel operational capability of GPU；It needs to complete in GPU
Have data cutout, interpolation, subtract direct current, FFT transform, modulus and take logarithm these calculating, they are rewritten into Kernel function,
The corresponding thread grid (Grid) of one kernel function, thread grid are made of a large amount of thread blocks (Block), thread block by
A large amount of thread (Thread) buildings；After the completion of thread layout, each data element has unique corresponding thread and thread block index,
A large amount of data element is all performed simultaneously identical program, the speed of data processing can be greatly improved, steps are as follows:
2.1) the collected original bin data of OCT is read on CPU to memory.
2.2) video memory is distributed on GPU using the cudaMalloc function in CUDA, since the present invention is by multiple B-
The SCAN successively parallel processing in GPU, therefore the video memory size distributed is determined by the data volume of a B-SCAN.
2.3) data copy ready on CPU to video memory is copied every time using the cudaMemcpy function in CUDA
Data volume size be a B-SCAN data volume.
2.4) call kernel function on CPU, parallel computation executed on GPU, complete data cutout, interpolation, subtract direct current,
FFT transform, modulus and take Logarithmic calculation, wherein FFT arithmetic section is realized using the library CUFFT that CUDA is carried.
2.5) result in GPU video memory is copied to CPU using the cudaMemcpy function in CUDA, multiple B-SCAN's
Calculated result is stored in three-dimensional array, obtains an individual data items, is prepared for subsequent three-dimensional visualization.
In the step 3), the treatment process of improved light projecting algorithm are as follows: in each pixel of display screen
Place issues one along direction of visual lines and passes through the ray of 3 d data field, and several equidistant resamplings are chosen on ray
Point assigns each sampled point different color values and opacity value, every three sampled points according to the classification results to functional value
It carries out curve fitting to obtain a quadratic polynomial, obtains a void using quadratic polynomial interpolation between every two sampled point and adopt
Sampling point not only increases sample rate in this way, also as less true samples point participates in calculating, makes to read volume textures and three lines
Property interpolation time-consuming is shorter.The color value and opacity value of true samples point and empty sampled point on every ray are finally pressed one
Fixed ruled synthesis obtains the color value of corresponding pixel points on display screen, forms final image.If ith sample point
Color value is Cnow, opacity value αnow, into ith sample point before color value be Cin, opacity value αin,
Color value after ith sample point is Cout, opacity αout.The calculation formula of image co-registration is as follows:
Coutαout=Cinαin+Cnowαnow(-αin) (1)
αout=αin+αnow(-αin) (2)
After ray enters three-dimensional data field, opacity be will continue to accumulate, when accumulated value reaches 1, it is believed that
Current and later sampled point does not contribute the color value and opacity of imaging plane pixel, to skip subsequent sampling
Point terminates in advance calculating, improves computational efficiency.
Technical concept of the invention are as follows: obtain three-dimensional initial data using frequency domain OCT system scanning finger sample, the data
The information such as Epidermal Fingerprint, sweat gland and skin corium fingerprint are contained, data volume is big and is not easy to analyze.It is transported by the high-speed parallel of GPU
Calculation ability accelerates the treatment process of three-dimensional initial data, realizes the real-time processing of OCT finger tip data.In order to further
Analysis, in conjunction with improved light projecting algorithm, to treated, data carry out three-dimensional visualization, and it is subcutaneous to clearly illustrate finger tip
The shape feature and spatial distribution of tissue.
Compared with prior art, the beneficial effects of the present invention are: compared with traditional two-dimentional fingerprint image acquisition, OCT
In addition to acquiring Epidermal Fingerprint, subcutaneous sweat gland, these new biological informations of fingerprint in skin corium are also acquired, is facilitated more smart
True fingerprint recognition；The treatment process of OCT finger tip data is accelerated by the high-speed parallel operational capability of GPU, Neng Goushi
It is handled when real；By improved light projecting algorithm, to treated, finger tip data carry out three-dimensional visualization, can be more straight
See the spatial relationship of the spatial position for accurately showing subcutaneous tissue of finger, size, geometry and it and surrounding tissue.
Detailed description of the invention
OCT system principle structural schematic diagram used in Fig. 1 present invention, wherein 1 is scanning galvanometer system, and 2 be plane mirror, 3
It is collimating mirror；4 be fiber coupler, and 5 be spectrometer.
Original bin data process flow diagram in Fig. 2 present invention.
Specific embodiment
The invention will be further described with embodiment with reference to the accompanying drawing:
Referring to Figures 1 and 2, a kind of real-time processing of OCT finger tip data and three-dimensional visualization method, include the following steps:
1) three-dimensional original bin data is obtained using frequency domain OCT system scanning finger sample；
2) according to interference imaging principle, parsing imaging is carried out to initial data, by the high-speed parallel operational capability pair of GPU
Treatment process is accelerated, and realizes the real-time processing of data and data are stored in three-dimensional array by treated；
3) using improved light projecting algorithm, to treated, finger tip data carry out three-dimensional visualization, show the table of finger tip
Fingerprint in skin fingerprint, sweat gland and skin corium.
Further, in the step 1), the process of frequency domain OCT system acquisition finger tip data: wideband light source SLD is issued low
Coherent light (central wavelength lambda 848nm, bandwidth Delta lambda 46nm) is divided into two-beam through four end fiber couplers, with 50/50 ratio
Example respectively enters reference arm and sample arm.Into the light in reference arm after collimation lens collimates, it is incident on is all-trans in parallel
It penetrates on mirror, and is reflected in parallel as reference light.It is focused on sample into the light in sample arm, passes through scanning galvanometer
System is scanned finger sample, is reflected as signal light by the scattering particles of different depth in finger sample.From reference arm
Converge in fiber coupler with the reference light and signal light returned in sample arm, superposition and interference occurs, interference signal is from coupling
The outgoing of the clutch other end enters spectrometer and is parsed, and is acquired by line array CCD, is sent into computer and is carried out by data collecting card
A series of data processing, then by treated, data carry out 2D or 3D imaging, referring to Fig. 1.The acquisition finger tip data
Steps are as follows:
1.1) an axial scan A-SCAN is carried out to a point on finger tip, it includes depth information that CCD, which collects one,
Line spectrum；
1.2) transversal scanning B-SCAN is carried out to finger tip, obtains the depth information spectrum of a plurality of A-SCAN composition；
1.3) the individual scan data of transversal scanning available one is successively repeatedly carried out backward.
The axial resolution of frequency domain OCT system used in the present invention is 7 μm, and lateral resolution is 11 μm, the pixel of CCD
It pair is 2048, then the number of pixels that an A-SCAN is obtained is 2048, and B-SCAN includes 500 A-SCAN signals, i.e.,
Finger sample a direction has carried out the acquisition of 500 depth datas, and sweep speed is 36klines/s, each B-SCAN size
For 500lines × 2048pixels/line × 2bytes/pixel, amount to 1.9MB data, an individual data items are by many B-
SCAN composition, data volume are very big.
In the step 2), data handling procedure includes being averaging to a plurality of A-SCAN, and every A-SCAN is counted respectively
According to interception, interpolation, subtracts direct current, FFT transform, modulus and take these operations of logarithm；Other than being averaging, other operations are all one
What mono- A-SCAN of A-SCAN was individually carried out, a large amount of single data elements can be calculated with complete independently, be suitble on GPU parallel
Processing, therefore the acceleration to data handling procedure can be realized by the high-speed parallel operational capability of GPU；It needs to complete in GPU
Have data cutout, interpolation, subtract direct current, FFT transform, modulus and take logarithm these calculating, they are rewritten into Kernel function,
The corresponding thread grid (Grid) of one kernel function, thread grid are made of a large amount of thread blocks (Block), thread block by
A large amount of thread (Thread) buildings；After the completion of thread layout, each data element has unique corresponding thread and thread block index,
A large amount of data element is all performed simultaneously identical program, the speed of data processing can be greatly improved, steps are as follows:
2.1) the collected original bin data of OCT is read on CPU to memory.
2.2) video memory is distributed on GPU using the cudaMalloc function in CUDA, since the present invention is by multiple B-
The SCAN successively parallel processing in GPU, therefore the video memory size distributed is determined by the data volume of a B-SCAN.
2.3) data copy ready on CPU to video memory is copied every time using the cudaMemcpy function in CUDA
Data volume size be a B-SCAN data volume.
2.4) call kernel function on CPU, parallel computation executed on GPU, complete data cutout, interpolation, subtract direct current,
FFT transform, modulus and take Logarithmic calculation, wherein FFT arithmetic section is realized using the library CUFFT that CUDA is carried.
2.5) result in GPU video memory is copied to CPU using the cudaMemcpy function in CUDA, multiple B-SCAN's
Calculated result is stored in three-dimensional array, obtains an individual data items, is prepared for subsequent three-dimensional visualization.
In the step 3), the treatment process of improved light projecting algorithm are as follows: in each pixel of display screen
Place issues one along direction of visual lines and passes through the ray of 3 d data field, and several equidistant resamplings are chosen on ray
Point assigns each sampled point different color values and opacity value, every three sampled points according to the classification results to functional value
It carries out curve fitting to obtain a quadratic polynomial, obtains a void using quadratic polynomial interpolation between every two sampled point and adopt
Sampling point not only increases sample rate in this way, also as less true samples point participates in calculating, makes to read volume textures and three lines
Property interpolation time-consuming is shorter.The color value and opacity value of true samples point and empty sampled point on every ray are finally pressed one
Fixed ruled synthesis obtains the color value of corresponding pixel points on display screen, forms final image.If ith sample point
Color value is Cnow, opacity value αnow, into ith sample point before color value be Cin, opacity value αin,
Color value after ith sample point is Cout, opacity αout.The calculation formula of image co-registration is as follows:
Coutαout=Cinαin+Cnowαnow(-αin) (1)
αout=αin+αnow(-αin) (2)
After ray enters three-dimensional data field, opacity be will continue to accumulate, when accumulated value reaches 1, it is believed that
Current and later sampled point does not contribute the color value and opacity of imaging plane pixel, to skip subsequent sampling
Point terminates in advance calculating, improves computational efficiency.
It is scanned respectively using finger tip of the frequency domain OCT system to living body finger tip and with artificial hand fingerstall, scan area 5mm
×5mm.Experimental development environment is Visual Studio 2010, and algorithm, which is realized, uses GLSL.When constructing transfer function, by epidermis
Fingerprint is set as brown, and subcutaneous tissue is set as green, to distinguish the different parts of finger tip.In three-dimensional visualization effect picture
Other than it can see the lines of finger tip Epidermal Fingerprint, it is further seen that the distribution of fingerprint in the subcutaneous sweat gland of finger tip, skin corium, it is false
Finger-stall surface and living body finger print are much like, but subcutaneously without sweat gland, then down again indistinctly it can be seen that living body finger print, this with just
Normal skin texture is not inconsistent completely, it was demonstrated that wide application of the OCT technology in fingerprint characteristic acquisition and fingerprint anti-counterfeit field
Prospect.
System of the invention is realized in surface chart, including render window, transfer function window and interactive window, render window
Show drawing result, transfer function window is used to be arranged color and opacity, and interactive window controls the rotation in tri- directions X, Y, Z
Turn.Button LoadBinData loads initial three-dimensional bin data, and completes the real-time processing of data, button LoadColorTF,
LoadAlphaTF loads color transmission function and opacity transfer function, realizes finger tip by improved light projecting algorithm
The three-dimensional visualization of data.
Claims (6)
1. a kind of real-time processing of OCT finger tip data and three-dimensional visualization method, which is characterized in that the method includes walking as follows
It is rapid:
1) three-dimensional original bin data is obtained using frequency domain OCT system scanning finger sample；
2) according to interference imaging principle, parsing imaging is carried out to initial data, by the high-speed parallel operational capability of GPU to processing
Process is accelerated, and realizes the real-time processing of data and data are stored in three-dimensional array by treated；
3) using improved light projecting algorithm, to treated, finger tip data carry out three-dimensional visualization, show that the epidermis of finger tip refers to
Fingerprint in line, sweat gland and skin corium.
2. real-time processing and the three-dimensional visualization method of a kind of OCT finger tip data as described in claim 1, which is characterized in that
In the step 1), the process of the frequency domain OCT system acquisition finger tip data: the low-coherent light warp that wideband light source SLD is issued
Four end fiber couplers are divided into two-beam, and the central wavelength lambda of low-coherent light is 848nm, bandwidth Delta lambda 46nm, with 50/50 ratio
Example respectively enters reference arm and sample arm；Into the light in reference arm after collimation lens collimates, it is incident on is all-trans in parallel
It penetrates on mirror, and is reflected in parallel as reference light；It is focused on sample into the light in sample arm, passes through scanning galvanometer
System is scanned finger sample, is reflected as signal light by the scattering particles of different depth in finger sample；From reference arm
Converge in fiber coupler with the reference light and signal light returned in sample arm, superposition and interference occurs, interference signal is from coupling
The outgoing of the clutch other end enters spectrometer and is parsed, and is acquired by line array CCD, is sent into computer and is carried out by data collecting card
A series of data processing, then by treated, data carry out 2D or 3D imaging.
3. real-time processing and the three-dimensional visualization method of a kind of OCT finger tip data as claimed in claim 2, which is characterized in that
In the step 1), the step of acquisition finger tip data, is as follows:
1.1) axial scan an A-SCAN, CCD are carried out to a point on finger tip and collects a line including depth information
Property spectrum；
1.2) transversal scanning B-SCAN is carried out to finger tip, obtains the depth information spectrum of a plurality of A-SCAN composition；
1.3) the individual scan data of transversal scanning available one is successively repeatedly carried out backward.
4. real-time processing and the three-dimensional visualization method of a kind of OCT finger tip data as described in one of claims 1 to 3, special
Sign is, in the step 2), data handling procedure includes being averaging to a plurality of A-SCAN, and every A-SCAN carries out data respectively
Interception, interpolation subtract direct current, FFT transform, modulus and take these operations of logarithm；Other than being averaging, other operations are all one
Mono- A-SCAN of A-SCAN is individually carried out, and a large amount of single data elements can be calculated with complete independently, is suitble to locate parallel on GPU
Reason, therefore the acceleration to data handling procedure is realized by the high-speed parallel operational capability of GPU；Need to complete in GPU has number
According to interception, interpolation, subtract direct current, FFT transform, modulus and take these calculating of logarithm, they are rewritten into Kernel function, one
Kernel function corresponds to a thread grid Grid, and thread grid is made of a large amount of thread block Block, and thread block is by a large amount of threads
Thread building；After the completion of thread layout, each data element has unique corresponding thread and thread block index, a large amount of data
Member is all performed simultaneously identical program.
5. real-time processing and the three-dimensional visualization method of a kind of OCT finger tip data as claimed in claim 4, which is characterized in that
The step of step 2), is as follows:
2.1) the collected original bin data of OCT is read on CPU to memory.
2.2) video memory is distributed on GPU using the cudaMalloc function in CUDA, due to the present invention be by multiple B-SCAN according to
The secondary parallel processing in GPU, therefore the video memory size distributed is determined by the data volume of a B-SCAN；
2.3) using the cudaMemcpy function in CUDA by data copy ready on CPU to video memory, the number copied every time
The data volume for being a B-SCAN according to amount size；
2.4) kernel function is called on CPU, parallel computation is executed on GPU, is completed data cutout, interpolation, is subtracted direct current, FFT change
It changes, modulus and take Logarithmic calculation, wherein FFT arithmetic section is realized using the library CUFFT that CUDA is carried；
2.5) result in GPU video memory is copied to CPU, the calculating of multiple B-SCAN using the cudaMemcpy function in CUDA
As a result it is stored in three-dimensional array, obtains an individual data items, prepare for subsequent three-dimensional visualization.
6. real-time processing and the three-dimensional visualization method of a kind of OCT finger tip data as described in one of claims 1 to 3, special
Sign is, in the step 3), the treatment process of improved light projecting algorithm are as follows: at each pixel of display screen
One is issued along direction of visual lines and passes through the ray of 3 d data field, and several equidistant resampling points are chosen on ray,
Assign each sampled point different color values and opacity value according to the classification results to functional value, every three sampled points carry out
Curve matching obtains a quadratic polynomial, obtains a void using quadratic polynomial interpolation between every two sampled point and samples
Point；Finally the color value of true samples point and empty sampled point on every ray and opacity value are closed by certain rule
At, obtain display screen on corresponding pixel points color value, form final image.If the color value of ith sample point is
Cnow, opacity value αnow, into ith sample point before color value be Cin, opacity value αin, by i-th
Color value after sampled point is Cout, opacity αout, the calculation formula of image co-registration is as follows:
Coutαout=Cinαin+Cnowαnow(-αin) (1)
αout=αin+αnow(-αin) (2)
After ray enters three-dimensional data field, opacity be will continue to accumulate, when accumulated value reaches 1, it is believed that it is current and with
Sampled point afterwards does not contribute the color value and opacity of imaging plane pixel, ties in advance to skip subsequent sampled point
Beam calculates.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811145157.8A CN109377549A (en) | 2018-09-29 | 2018-09-29 | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811145157.8A CN109377549A (en) | 2018-09-29 | 2018-09-29 | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109377549A true CN109377549A (en) | 2019-02-22 |
Family
ID=65402527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811145157.8A Pending CN109377549A (en) | 2018-09-29 | 2018-09-29 | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109377549A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110119699A (en) * | 2019-04-28 | 2019-08-13 | 清华大学深圳研究生院 | Interior fingerprint extraction method, device, system and storage medium and fingerprint identification method |
| CN110309699A (en) * | 2019-03-22 | 2019-10-08 | 浙江工业大学 | A kind of subcutaneous pore figure extraction method based on OCT |
| CN110309738A (en) * | 2019-06-17 | 2019-10-08 | 深圳大学 | The method that a kind of pair of OCT fingerprint image is labeled |
| CN111951378A (en) * | 2020-07-31 | 2020-11-17 | 中国电子科技集团公司第四十一研究所 | Visual view generation method for multi-source electromagnetic data |
| CN111986135A (en) * | 2020-08-26 | 2020-11-24 | 成都蕊感微电子有限公司 | Image anti-counterfeiting identification method, device, equipment and computer readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101976458A (en) * | 2010-10-15 | 2011-02-16 | 华中科技大学 | Medical image three-dimensional visualization method applied to mobile handheld equipment |
| CN102613960A (en) * | 2012-04-16 | 2012-08-01 | 北京信息科技大学 | Method for rectifying position and phase of frequency-domain optical coherence tomography signal |
| CN103106685A (en) * | 2013-01-16 | 2013-05-15 | 东北大学 | Abdominal viscera three dimension visualization method based on graphic processing unit (GPU) |
| US20140241596A1 (en) * | 2013-02-28 | 2014-08-28 | The Regents Of The University Of California | Apparatus and Method for Capturing a Vital Vascular Fingerprint |
| CN107146262A (en) * | 2017-04-18 | 2017-09-08 | 清华大学深圳研究生院 | The three-dimensional visualization method and system of a kind of OCT image |
| CN107479831A (en) * | 2017-08-11 | 2017-12-15 | 浙江工业大学 | A kind of OCT volume data method for carrying based on Zynq platforms |
-
2018
- 2018-09-29 CN CN201811145157.8A patent/CN109377549A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101976458A (en) * | 2010-10-15 | 2011-02-16 | 华中科技大学 | Medical image three-dimensional visualization method applied to mobile handheld equipment |
| CN102613960A (en) * | 2012-04-16 | 2012-08-01 | 北京信息科技大学 | Method for rectifying position and phase of frequency-domain optical coherence tomography signal |
| CN103106685A (en) * | 2013-01-16 | 2013-05-15 | 东北大学 | Abdominal viscera three dimension visualization method based on graphic processing unit (GPU) |
| US20140241596A1 (en) * | 2013-02-28 | 2014-08-28 | The Regents Of The University Of California | Apparatus and Method for Capturing a Vital Vascular Fingerprint |
| CN107146262A (en) * | 2017-04-18 | 2017-09-08 | 清华大学深圳研究生院 | The three-dimensional visualization method and system of a kind of OCT image |
| CN107479831A (en) * | 2017-08-11 | 2017-12-15 | 浙江工业大学 | A kind of OCT volume data method for carrying based on Zynq platforms |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309699A (en) * | 2019-03-22 | 2019-10-08 | 浙江工业大学 | A kind of subcutaneous pore figure extraction method based on OCT |
| CN110309699B (en) * | 2019-03-22 | 2021-06-18 | 浙江工业大学 | Automatic extraction method of subcutaneous sweat pore map based on OCT |
| CN110119699A (en) * | 2019-04-28 | 2019-08-13 | 清华大学深圳研究生院 | Interior fingerprint extraction method, device, system and storage medium and fingerprint identification method |
| CN110309738A (en) * | 2019-06-17 | 2019-10-08 | 深圳大学 | The method that a kind of pair of OCT fingerprint image is labeled |
| CN110309738B (en) * | 2019-06-17 | 2022-09-30 | 深圳大学 | Method for labeling OCT fingerprint image |
| CN111951378A (en) * | 2020-07-31 | 2020-11-17 | 中国电子科技集团公司第四十一研究所 | Visual view generation method for multi-source electromagnetic data |
| CN111951378B (en) * | 2020-07-31 | 2023-11-21 | 中国电子科技集团公司第四十一研究所 | Visual view generation method of multi-source electromagnetic data |
| CN111986135A (en) * | 2020-08-26 | 2020-11-24 | 成都蕊感微电子有限公司 | Image anti-counterfeiting identification method, device, equipment and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109377549A (en) | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method | |
| CN101513366B (en) | PS-OCT-based three dimension visualized imaging method | |
| CN107788950B (en) | Blood flow imaging method and system based on self-adaptive threshold segmentation | |
| US9693728B2 (en) | Systems and methods for measuring mechanical properties of deformable materials | |
| CN107924580A (en) | The visualization of surface volume mixing module in medical imaging | |
| KR102580984B1 (en) | Image processing method, program and recording medium | |
| CN101264002A (en) | Three-dimensional endoscopic measurement device and method based on grating projection | |
| US9008407B2 (en) | Noise reduction processing method and apparatus for a biological tissue image | |
| Gao et al. | Medical images classification for skin cancer using quantitative image features with optical coherence tomography | |
| CN106455978A (en) | Diagnostic device for dermatology with merged oct and epiluminescence dermoscopy | |
| CN112057049B (en) | Optical coherent blood flow radiography method and system based on multi-dimensional feature space | |
| Sousedik et al. | Volumetric fingerprint data analysis using optical coherence tomography | |
| JP5847454B2 (en) | Subject information acquisition apparatus, display control method, and program | |
| CN108694726A (en) | medical image three-dimensional reconstruction system | |
| Yadav et al. | A study on automatic early detection of skin cancer | |
| US20230200930A1 (en) | Intelligent Surgical Marker | |
| CN113706567B (en) | Blood flow imaging quantification processing method and device combined with blood vessel morphological characteristics | |
| Nakao et al. | Visualizing in vivo brain neural structures using volume rendered feature spaces | |
| CN210810980U (en) | Skin imaging equipment in ultra-wide range | |
| CN113362412A (en) | Speckle spectrum information reconstruction method and device based on deep learning | |
| Kwon et al. | GPU-accelerated 3D mipmap for real-time visualization of ultrasound volume data | |
| JP2016027924A (en) | Subject information acquisition apparatus and display method | |
| JP2020144868A (en) | Volume surface generator using wave vectors | |
| Miao et al. | Noise reduction of fingertip OCT image based on generated unpaired high-quality datasets | |
| JP2018192309A (en) | Information processing device, information processing method, and program |
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: 20190222 |
|
| RJ01 | Rejection of invention patent application after publication |