CN110309699A - A kind of subcutaneous pore figure extraction method based on OCT - Google Patents
A kind of subcutaneous pore figure extraction method based on OCT Download PDFInfo
- Publication number
- CN110309699A CN110309699A CN201910219874.9A CN201910219874A CN110309699A CN 110309699 A CN110309699 A CN 110309699A CN 201910219874 A CN201910219874 A CN 201910219874A CN 110309699 A CN110309699 A CN 110309699A
- Authority
- CN
- China
- Prior art keywords
- point
- image
- cuticula
- oct
- profile
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
- G06V40/1353—Extracting features related to minutiae or pores
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
A kind of subcutaneous pore figure extracting method based on OCT includes the following steps: 1) to carry out gray value calculus of differences to each pixel of every OCT image, and selection result is greater than the point of threshold value as initial characteristics point set；2) Hough transformation is applied, cuticula feature point set is separated from initial characteristics point concentration, and carries out quadratic polynomial to it and is fitted to obtain cuticula profile, while removing the characteristic point for being located at cutin profile near its circumference and top；3) characteristic point outside papillaris pars profile is removed from the distant to the near, obtains accurate papillaris pars feature point set, and is carried out cubic interpolation to it and be fitted to obtain papillaris pars profile；4) sweat gland tangent line is obtained according to the position of two profiles, it is the subcutaneous pore figure of W × N that the sweat gland tangent line obtained in all OCT images, which is then spliced into size, obtains final result using image enhancement.The present invention can obtain correct sweat gland tangent line, finally obtain clearly subcutaneous pore figure.
Description
Technical field
The present invention relates to fingerprint recognition field, in particular to a kind of subcutaneous pore figure side of automatically extracting based on OCT system
Method.
Background technique
Fingerprint identification technology has become current individual identification or recognizes since it is unique, permanent and facilitates collection
Most commonly used biometric feature in card.Currently, be distributed in the pore between fingerprint ridge line, as the 3 of fingerprint grade feature, just
It is applied in fingerprint recognition field more and more.However, when finger surface there are dirt, sweat and scar or is cut
When the damage of mouth bring unrepairable, pore just will receive destruction, be unable to complete identification mission.In addition, epidermis pore has
Two states are opened and closed, also will affect the accuracy of identification.
Studies have shown that the supracutaneous pore of finger, is formed by subcutaneous sweat gland.Sweat gland is grown on the friendship of corium and epidermis
Papillaris pars (papilla) at boundary, extends to skin surface, to form pore.It can be seen that subcutaneous group of finger
It knits and carries out crosscutting obtained sweat gland cross section (i.e. subcutaneous pore figure) compared with epidermis pore, there is not easily damaged, complete stabilization
Advantage.At the same time, optical coherence tomography (optical coherence tomography, OCT) this non-intruding
Property imaging technique, the information of 1~3mm depth, obtains the 3D volume data of finger print, this is under available human skin surface
It obtains the subcutaneous pore figure of high-resolution and provides possibility.
Summary of the invention
Accuracy in order to overcome the shortcomings of existing subcutaneous pore figure extracting mode is poor, and the invention proposes one kind to be based on
The subcutaneous pore figure extraction method of OCT, obtains the cuticula of finger and the outline position of papillaris pars, thus according to two first
The position of a profile determines the crosscutting position of sweat gland, realizes and extracts the higher universality of subcutaneous pore figure and robustness.
To achieve the goals above, the technical solution adopted by the present invention are as follows:
A kind of subcutaneous pore figure extraction method based on OCT, includes the following steps:
1) OCT fingerprint volume data size is set as W × H × N, i.e., is made of, indicates the OCT image that N resolution ratio are W × H
The spatially vertical tangent plane of continuous N finger prints carries out gray value calculus of differences to each pixel of every OCT image, and
Selection result is greater than the point of threshold value as initial characteristics point set；
2) since cuticula characteristic point almost forms the straight line of a slight curvature, using Hough transformation, by cuticula spy
Sign point set is separated from initial characteristics point concentration, and is carried out quadratic polynomial to it and be fitted to obtain cuticula profile, and in point
Removal is concentrated to be located at the characteristic point of cutin profile near its circumference and top；
3) characteristic point outside papillaris pars profile is removed from the distant to the near, obtains accurate papillaris pars feature point set, and to it
Cubic interpolation is carried out to be fitted to obtain papillaris pars profile；
4) sweat gland tangent line (being made of w pixel) is obtained according to the position of two profiles, then will be obtained in all OCT images
It is the subcutaneous pore figure of W × N that the sweat gland tangent line obtained, which is spliced into size, obtains final result using image enhancement.
Further, in the step 1), due to big in cuticula and papillaris pars gray-value variation, gray value differences partite transport is used
It calculates to extract the characteristic point of OCT image:
Wherein, g (x, y) is expressed as the gray value of coordinate points (x, y), and 0≤x < W, 0≤y < H-1, y value is bigger, and representative is deeper
Position, i.e. image gets over lower position, meets all the points in above formula as initial characteristics point set P0。
Further, the step 2) includes the following steps:
2.1) larger in the gray-value variation of cuticula and papillaris pars due to image, initial characteristics point concentration is distributed in cutin
On layer and papillaris pars；Meanwhile cuticula characteristic point almost forms the straight line of a slight curvature, therefore is mentioned using Hough transformation
Take cuticula.Initial characteristics point image is converted into bianry image first, then to bianry image application Hough transformation, in result square
Hn=6 maximum value is found in battle array, obtains 6 line segments being distributed in cuticula characteristic point；By the gap between these line segments
It connects, simultaneously for the two lines section of the image leftmost side and the rightmost side, extends to image border to the left and to the right respectively, this
Sample is just at a continuous line segment L；
2.2) because line segment L is overlapped with cuticula substantially, to any one point j, angle is grouped into as long as meeting following formula
In matter layer characteristic point:
Dis (j, L) <=v (j ∈ P0) (2)
That is point j is less than threshold value v=1 at a distance from straight line L, obtains cuticula feature point set P in this waySC, while to PSCIt carries out
Quadratic polynomial is fitted to obtain cuticula profile LSC；
2.3) removal is located at LSCTop and all and LSCCharacteristic point of the distance less than 10, obtains for extracting papillaris pars
The feature point set P ' of characteristic pointPL。
Further, the step 3) includes the following steps:
3.1) removal first is apart from the farther away characteristic point of papillaris pars: by point set P 'PLThe middle all the points i for meeting following formula is gone
It removes:
Wherein, W is the width of image, (xi,yi) be point i coordinate；Q is using least square method by point set P 'PLFitting
Obtained approximate conic section,Q is represented in the corresponding ordinate value in point i abscissa place；T1For distance threshold, 25 are taken；
3.2) be apart from the closer characteristic point to be removed of papillaris pars sweat gland level of approximation linear segments, be located at papillaris pars on
Side；Since papillaris pars profile is the curve of continuously smooth variation, the slope of two endpoints of sweat gland segment is larger and symbol is on the contrary, two
The distance between a endpoint is smaller and segment is distributed in level of approximation, therefore close to two endpoints of sweat gland of papillaris pars is defined as:
Wherein, (XA,YA) and (XB,YB) respectively indicate A endpoint and B endpoint, G(A)And G(B)Respectively indicate two endpoints of A, B
Slope, T2=10, Chinese gland model parameter t1Take the diameter length slightly larger than sweat gland, t2Take the value of very little, here, take t1=20
And t2=4, the characteristic point between the terminal A of above formula and B will be met and removed, that retain is then papillaris pars feature point set PPL, make
With cubic spline interpolation, by papillaris pars feature point set PPLIt is fitted to smooth continuous papillaris pars profile LPL。
The step 4) includes the following steps:
4.1) sweat gland tangent line L is obtained with following formulaSG:
(LSG)x=(0.3* (LSC)x+0.7*(LPL)x) (5)
That is LSGBetween two contour lines, but closer to papillaris pars profile, by LSGPixel on position is taken out, and obtains
To the straight line for an a length of W, then according to image sequence, complete subcutaneous pore image at face is spliced by line；
4.2) final subcutaneous pore figure is obtained using image enhancement to the subcutaneous pore image of acquisition.
Compared with prior art, beneficial effects of the present invention are shown: can be stablized for different nipple layer depths
Cuticula and papillaris pars characteristic point classification and extraction to obtain correct sweat gland tangent line finally obtain clearly subcutaneous sweat
Hole pattern.
Detailed description of the invention
Fig. 1 is the flow chart of inventive algorithm；
Fig. 2 is OCT image；
Fig. 3 is initial characteristics point diagram；
Fig. 4 is the result of Hough transformation；
Fig. 5 is the result that the line segment for obtaining Hough transformation connects；
Fig. 6 is cuticula profile；
Fig. 7 is remaining papillaris pars characteristic point point set P 'PL；
Fig. 8 is papillaris pars profile；
What the medium line in Fig. 9 represented is sweat gland tangent line；
Figure 10 is the subcutaneous pore figure finally obtained.
Specific embodiment
The invention will be further described with embodiment with reference to the accompanying drawing:
A kind of referring to Fig.1~Figure 10, subcutaneous pore figure extraction method based on OCT, includes the following steps:
1) OCT fingerprint volume data size is set to be made of the OCT image that resolution ratio is, continuous N on representation space
The vertical tangent plane for opening finger print carries out gray value calculus of differences to each pixel of every OCT image, and selection result is greater than threshold
The point of value is as initial characteristics point set；
Due to big in cuticula and papillaris pars gray-value variation, the spy of OCT image is extracted using gray value calculus of differences
Sign point:
D (x, y)=g (x, y+1)-g (x, y)
D (x, y) > 5
Wherein, g (x, y) is expressed as the gray value (0≤x < W, 0≤y < H-1) of coordinate points (x, y), and the bigger representative of y value is more
Deep position (i.e. image gets over lower position), meets all the points in above formula as initial characteristics point set P0；
2) since cuticula characteristic point almost forms the straight line of a slight curvature, using Hough transformation, by cuticula spy
Sign point set is separated from initial characteristics point concentration, and is carried out quadratic polynomial to it and be fitted to obtain cuticula profile, and in point
Removal is concentrated to be located at the characteristic point of cutin profile near its circumference and top；Include the following steps:
2.1) larger in the gray-value variation of cuticula and papillaris pars due to image, initial characteristics point concentration is distributed in cutin
On layer and papillaris pars；Meanwhile cuticula characteristic point almost forms the straight line of a slight curvature, therefore is mentioned using Hough transformation
Take cuticula.Initial characteristics point image is converted into bianry image first, then to bianry image application Hough transformation, in result square
Hn=6 maximum value is found in battle array, obtains 6 line segments being distributed in cuticula characteristic point；By the gap between these line segments
It connects, simultaneously for the two lines section of the image leftmost side and the rightmost side, extends to image border to the left and to the right respectively, this
Sample is just at a continuous line segment L；
2.2) because line segment L is overlapped with cuticula substantially, to any one point j, angle is grouped into as long as meeting following formula
In matter layer characteristic point:
Dis (j, L) <=v (j ∈ P0)
That is point j is less than threshold value v=1 at a distance from straight line L, obtains cuticula feature point set P in this waySC, while to PSCIt carries out
Quadratic polynomial is fitted to obtain cuticula profile LSC。
2.3) removal is located at LSCTop and all and LSCCharacteristic point of the distance less than 10, obtains for extracting papillaris pars
The feature point set P ' of characteristic pointPL；
3) characteristic point outside papillaris pars profile is removed from the distant to the near, obtains accurate papillaris pars feature point set, and to it
Cubic interpolation is carried out to be fitted to obtain papillaris pars profile；Include the following steps:
3.1) removal first is apart from the farther away characteristic point of papillaris pars: by point set P 'PLThe middle all the points i for meeting following formula is gone
It removes:
Wherein, W is the width of image, (xi,yi) be point i coordinate；Q is using least square method by point set P 'PLFitting
Obtained approximate conic section,Q is represented in the corresponding ordinate value in point i abscissa place；T1For distance threshold, 25 are taken；
3.2) be apart from the closer characteristic point to be removed of papillaris pars sweat gland level of approximation linear segments, be located at papillaris pars on
Side.Since papillaris pars profile is the curve of continuously smooth variation, the slope of two endpoints of sweat gland segment is larger and symbol is on the contrary, two
The distance between a endpoint is smaller and segment is distributed in level of approximation.Therefore close to two endpoints of sweat gland of papillaris pars is defined as:
Wherein, (XA,YA) and (XB,YB) respectively indicate A endpoint and B endpoint, G(A)And G(B)Respectively indicate two endpoints of A, B
Slope, T2=10, Chinese gland model parameter t1Take the diameter length slightly larger than sweat gland, t2Take the value of very little, here, take t1=20
And t2=4, the characteristic point between the terminal A of above formula and B will be met and removed, that retain is then papillaris pars feature point set PPL, make
With cubic spline interpolation, by papillaris pars feature point set PPLIt is fitted to smooth continuous papillaris pars profile LPL；
4) papillaris pars profile being obtained using interpolation, sweat gland tangent line is obtained (by w pixel group according to the position of two profiles
At), it is the subcutaneous pore figure of W × N that the sweat gland tangent line obtained in all OCT images, which is then spliced into size, is increased using image
It is strong to obtain final result；Include the following steps:
4.1) sweat gland tangent line L is obtained with following formulaSG:
(LSG)x=(0.3* (LSC)x+0.7*(LPL)x)
That is LSGBetween two contour lines, but closer to papillaris pars profile.By LSGPixel on position is taken out, and obtains
To the straight line for an a length of W, then according to image sequence, complete subcutaneous pore image at face is spliced by line；
4.2) final subcutaneous pore figure is obtained using image enhancement to the subcutaneous pore image of acquisition.
Claims (5)
1. a kind of subcutaneous pore figure extraction method based on OCT, which is characterized in that the described method comprises the following steps:
1) OCT fingerprint volume data size is set as W × H × N, i.e., is made of the OCT image that N resolution ratio are W × H, representation space
The vertical tangent plane of upper continuous N finger prints carries out gray value calculus of differences to each pixel of every OCT image, and selects
As a result greater than the point of threshold value as initial characteristics point set；
2) since cuticula characteristic point almost forms the straight line of a slight curvature, using Hough transformation, by cuticula characteristic point
Collection is separated from initial characteristics point concentration, and is carried out quadratic polynomial to it and be fitted to obtain cuticula profile, while removing position
Characteristic point in cutin profile near its circumference and top；
3) characteristic point outside papillaris pars profile is removed from the distant to the near, obtains accurate papillaris pars feature point set, and carry out to it
Cubic interpolation is fitted to obtain papillaris pars profile；
4) sweat gland tangent line is obtained according to the position of two profiles, the sweat gland tangent line is made of w pixel, then by all OCT
It is the subcutaneous pore figure of W × N that the sweat gland tangent line obtained in image, which is spliced into size, obtains final result using image enhancement.
2. a kind of subcutaneous pore figure extraction method based on OCT as described in claim 1, it is characterised in that: the step
It is rapid 1) in, due to big in cuticula and papillaris pars gray-value variation, the feature of OCT image is extracted using gray value calculus of differences
Point:
Wherein, g (x, y) is expressed as the gray value of coordinate points (x, y), and 0≤x < W, 0≤y < H-1, y value is bigger to represent deeper position
It sets, i.e., image gets over lower position, meets all the points in above formula as initial characteristics point set P0。
3. a kind of subcutaneous pore figure extraction method based on OCT as claimed in claim 1 or 2, it is characterised in that: described
In step 2), include the following steps:
2.1) larger in the gray-value variation of cuticula and papillaris pars due to image, initial characteristics point concentration be distributed in cuticula and
On papillaris pars, meanwhile, cuticula characteristic point almost forms the straight line of a slight curvature, therefore extracts angle using Hough transformation
Initial characteristics point image is converted to bianry image first by matter layer, then to bianry image application Hough transformation, in matrix of consequence
Hn=6 maximum value is found, 6 line segments being distributed in cuticula characteristic point are obtained, the gap between these line segments is connected
Get up, simultaneously for the two lines section of the image leftmost side and the rightmost side, extends to image border to the left and to the right respectively, thus
Row is at a continuous line segment L；
2.2) because line segment L is overlapped with cuticula substantially, to any one point j, cuticula is grouped into as long as meeting following formula
In characteristic point:
Dis (j, L) <=v (j ∈ P0) (2)
That is point j is less than threshold value v=1 at a distance from straight line L, obtains cuticula feature point set P in this waySCIt is special with being used to extract papillaris pars
Levy other feature point sets P' of pointPL, while to PSCQuadratic polynomial is carried out to be fitted to obtain cuticula profile LSC；
2.3) removal is located at LSCTop and all and LSCCharacteristic point of the distance less than 10.
4. a kind of subcutaneous pore figure extraction method based on OCT as claimed in claim 1 or 2, it is characterised in that: described
In step 3), include the following steps:
3.1) removal first is apart from the farther away characteristic point of papillaris pars: by point set P'PLThe middle all the points i removal for meeting following formula:
Wherein, W is the width of image, (xi,yi) be point i coordinate；Q is using least square method by point set P'PLWhat fitting obtained
Approximate conic section,Q is represented in the corresponding ordinate value in point i abscissa place；T1For distance threshold, 25 are taken；
3.2) be apart from the closer characteristic point to be removed of papillaris pars sweat gland level of approximation linear segments, be located at papillaris pars above,
Since papillaris pars profile is the curve of continuously smooth variation, the slope of two endpoints of sweat gland segment is larger and symbol is on the contrary, two
The distance between endpoint is smaller and segment is distributed in level of approximation, therefore close to two endpoints of sweat gland of papillaris pars is defined as:
Wherein, (XA,YA) and (XB,YB) respectively indicate A endpoint and B endpoint, G(A)And G(B)Respectively indicate the oblique of two endpoints of A, B
Rate, T2=10, Chinese gland model parameter t1Take the diameter length greater than sweat gland, t2Take the value of very little, here, take t1=20 and t2=4,
The removal of the characteristic point between the terminal A of above formula and B will be met, that retain is then papillaris pars feature point set PPL, use sample three times
Interpolation, by papillaris pars feature point set PPLIt is fitted to smooth continuous papillaris pars profile LPL。
5. a kind of subcutaneous pore figure extraction method based on OCT as claimed in claim 1 or 2, it is characterised in that: described
In step 4), include the following steps:
4.1) sweat gland tangent line L is obtained with following formulaSG:
(LSG)x=(0.3* (LSC)x+0.7*(LPL)x) (6)
That is LSGBetween two contour lines, but closer to papillaris pars profile, by LSGPixel on position is taken out, and obtains being one
The straight line of a length of W of item at face is spliced into complete subcutaneous pore image by line then according to image sequence；
4.2) final subcutaneous pore figure is obtained using image enhancement to the subcutaneous pore image of acquisition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910219874.9A CN110309699B (en) | 2019-03-22 | 2019-03-22 | Automatic extraction method of subcutaneous sweat pore map based on OCT |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910219874.9A CN110309699B (en) | 2019-03-22 | 2019-03-22 | Automatic extraction method of subcutaneous sweat pore map based on OCT |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110309699A true CN110309699A (en) | 2019-10-08 |
| CN110309699B CN110309699B (en) | 2021-06-18 |
Family
ID=68074343
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910219874.9A Active CN110309699B (en) | 2019-03-22 | 2019-03-22 | Automatic extraction method of subcutaneous sweat pore map based on OCT |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110309699B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112528728A (en) * | 2020-10-16 | 2021-03-19 | 深圳市银星智能科技股份有限公司 | Image processing method and device for visual navigation and mobile robot |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070165914A1 (en) * | 2003-10-23 | 2007-07-19 | Lumeniq, Inc. | Systems and methods relating to AFIS recognition, extraction, and 3-D analysis strategies |
| CN102846309A (en) * | 2011-05-23 | 2013-01-02 | 索尼公司 | Image processing device, image processing method, program, and recording medium |
| CN107563364A (en) * | 2017-10-23 | 2018-01-09 | 清华大学深圳研究生院 | The discriminating conduct of the fingerprint true and false and fingerprint identification method based on sweat gland |
| CN108446633A (en) * | 2018-03-20 | 2018-08-24 | 深圳大学 | A kind of method, system and device of novel finger print automatic anti-fake and In vivo detection |
| CN108932507A (en) * | 2018-08-06 | 2018-12-04 | 深圳大学 | A kind of automatic anti-fake method and system based on OCT fingerprint image |
| CN109247911A (en) * | 2018-07-25 | 2019-01-22 | 浙江工业大学 | A kind of multi-modal characteristic synchronization acquisition system of finger |
| CN109310337A (en) * | 2016-06-20 | 2019-02-05 | 公立大学法人大阪市立大学 | Skin diagnosis device, skin condition output method, program and recording medium |
| CN109377549A (en) * | 2018-09-29 | 2019-02-22 | 浙江工业大学 | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method |
-
2019
- 2019-03-22 CN CN201910219874.9A patent/CN110309699B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070165914A1 (en) * | 2003-10-23 | 2007-07-19 | Lumeniq, Inc. | Systems and methods relating to AFIS recognition, extraction, and 3-D analysis strategies |
| CN102846309A (en) * | 2011-05-23 | 2013-01-02 | 索尼公司 | Image processing device, image processing method, program, and recording medium |
| CN109310337A (en) * | 2016-06-20 | 2019-02-05 | 公立大学法人大阪市立大学 | Skin diagnosis device, skin condition output method, program and recording medium |
| CN107563364A (en) * | 2017-10-23 | 2018-01-09 | 清华大学深圳研究生院 | The discriminating conduct of the fingerprint true and false and fingerprint identification method based on sweat gland |
| CN108446633A (en) * | 2018-03-20 | 2018-08-24 | 深圳大学 | A kind of method, system and device of novel finger print automatic anti-fake and In vivo detection |
| CN109247911A (en) * | 2018-07-25 | 2019-01-22 | 浙江工业大学 | A kind of multi-modal characteristic synchronization acquisition system of finger |
| CN108932507A (en) * | 2018-08-06 | 2018-12-04 | 深圳大学 | A kind of automatic anti-fake method and system based on OCT fingerprint image |
| CN109377549A (en) * | 2018-09-29 | 2019-02-22 | 浙江工业大学 | A kind of real-time processing of OCT finger tip data and three-dimensional visualization method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112528728A (en) * | 2020-10-16 | 2021-03-19 | 深圳市银星智能科技股份有限公司 | Image processing method and device for visual navigation and mobile robot |
| CN112528728B (en) * | 2020-10-16 | 2024-03-29 | 深圳银星智能集团股份有限公司 | Image processing method and device for visual navigation and mobile robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110309699B (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105893949B (en) | A kind of method for detecting lane lines under complex road condition scene | |
| CN106295648B (en) | A kind of low quality file and picture binary coding method based on multi-optical spectrum imaging technology | |
| CN108694393A (en) | A kind of certificate image text area extraction method based on depth convolution | |
| CN109241973B (en) | Full-automatic soft segmentation method for characters under texture background | |
| CN1312625C (en) | Character extracting method from complecate background color image based on run-length adjacent map | |
| CN103870808A (en) | Finger vein identification method | |
| CN109376708B (en) | Method for extracting ROI | |
| CN107749049B (en) | Vein distribution display method and device | |
| CN107066952A (en) | A kind of method for detecting lane lines | |
| CN102663378A (en) | Method for indentifying joined-up handwritten characters | |
| CN108921813A (en) | Unmanned aerial vehicle detection bridge structure crack identification method based on machine vision | |
| CN110751029B (en) | Maximum curvature-based adaptive finger vein line extraction method | |
| CN116740054B (en) | Tongue image tooth trace detection method based on image processing | |
| CN115731257A (en) | Leaf form information extraction method based on image | |
| CN102136074B (en) | Man-machine interface (MMI) based wood image texture analyzing and identifying method | |
| CN109766850A (en) | Fingerprint image matching method based on Fusion Features | |
| WO2022121021A1 (en) | Identity card number detection method and apparatus, and readable storage medium and terminal | |
| CN108710883A (en) | A kind of complete conspicuousness object detecting method using contour detecting | |
| CN110309699A (en) | A kind of subcutaneous pore figure extraction method based on OCT | |
| CN115272306A (en) | Solar cell panel grid line enhancement method utilizing gradient operation | |
| Abdulla et al. | Off-Line Arabic Handwritten Word Segmentation Using Rotational Invariant Segments Features. | |
| CN105528795B (en) | A kind of infrared face dividing method using annular shortest path | |
| CN109460763B (en) | Text region extraction method based on multilevel text component positioning and growth | |
| CN112258518A (en) | Sea-sky-line extraction method and device | |
| CN112070684A (en) | Method for repairing nail carving words based on morphological prior characteristics |
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 |