CN101286194A - Fingerprint frame line filtering method - Google Patents
Fingerprint frame line filtering method Download PDFInfo
- Publication number
- CN101286194A CN101286194A CNA2007100179280A CN200710017928A CN101286194A CN 101286194 A CN101286194 A CN 101286194A CN A2007100179280 A CNA2007100179280 A CN A2007100179280A CN 200710017928 A CN200710017928 A CN 200710017928A CN 101286194 A CN101286194 A CN 101286194A
- Authority
- CN
- China
- Prior art keywords
- window
- row
- image
- frame line
- value
- 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
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The invention relates to a fingerprint image frame line deletion method, which pertains to the digital image processing and fingerprint identification technical fields. The filtering method is characterized in that: an image with the size of m multiplied by n is firstly carried out the band-stop filter pre-treatment, the image after the pre-treatment is carried out the binary processing, two windows with the size of m multiplied by 9 and 9 multiplied by n are used for scanning the whole image along the longitudinal direction and the transverse direction, the minimum value of the sums of rows of pixels is calculated within the window field with the size of m multiplied by 9, the minimum value of the sums of lines of pixels is calculated within the window field with the size of 9 multiplied by n, the filling of the image is carried out according to the values and the deletion of the image frame lines is completed. The deletion method has the advantages that: the deletion method is applicable to the removal of an interference which is generally existed in an automatic fingerprint identification system, thus allowing the automatic fingerprint identification to be more accurate.
Description
Technical field
The present invention relates to a kind of filtering method of fingerprint frame line, be used for removing a kind of ubiquitous interference, make that automatic fingerprint recognition is more accurate at Automated Fingerprint Identification System.
Background technology
Automated Fingerprint Identification System (Automated Fingerprint Identification System is called for short AFIS) with its security, reliability and high efficiency, is with a wide range of applications in fields such as public security, information security and ecommerce.Shown in Figure 1 promptly is the stream of a complete Automated Fingerprint Identification System.From figure we as can be seen Automated Fingerprint Identification System form by several parts such as image pre-service, feature extraction, classification and couplings.Fingerprint image will classify and mate through after the pretreated series of steps, and registration is the final goal of Automated Fingerprint Identification System just.The singular point classification is just arranged in the several different methods of fingerprint classification, and in fact this method is exactly to utilize the central point of fingerprint and the position and the quantitative relation of trigpoint that fingerprint is belonged in the taxonomic hierarchies.Matching process based on figure is also arranged in the finger print matching method, and mate the location that this method also is based on central point, and therefore as can be seen in the Automated Fingerprint Identification System, the central point location is an important link.
The central point positioning algorithm based has Poincar ' index algorithm, and directional diagram neighborhood sine ratio algorithm also has and follow the tracks of the track algorithm that obtains central point in refinement figure.We test a large amount of central point location algorithms, find that framing mask has very big influence to the result who handles, as shown in Figure 2 when fingerprint image is handled.
As can be seen from the figure, do not have the Core Point in Fingerprint location algorithm of frame line very accurate, and the central point location that influences that has the image of frame line to be subjected to frame fall flat.The existence of frame line does not meet the texture structure of fingerprint fully, and the relation that the pixel value of each point produces on the pixel value of each point and the fingerprint image on the frame line is similar to the radian value of central point, and the system mistake that is automatically recognized detects and is central point.
Summary of the invention
The technical matters that solves
For fear of the deficiencies in the prior art part, the present invention proposes a kind of filtering method of fingerprint frame line, can realize framing mask is handled, to eliminate the influence of frame to the output result.
Technical scheme
Thought of the present invention is: frame line is actual to be during from printing ink stamp fingerprint, and the printing sorting track on the capture card has been brought into database in scanning process.These lines have higher density, and are connected with the fingerprint line.But, on level and vertical direction relatively, be to run through entire image according to the frame line of fingerprint image, it is to take on a different character a little that this and fingerprint run through on the direction at one on the entire image.The gray difference of each fingerprint image itself is bigger in addition, and also difference to some extent of the pixel value of each point in the gray level image utilizes fingerprint image not have Calculation Method in the bianry image of feature loss, eliminates the frame line of fingerprint.
The concrete steps of the technology of the present invention feature are as follows:
1) at first to carrying out the bandreject filtering pre-service for the image of m * n for width of cloth size:
D in the formula
l, D
hBe cutoff frequency; (u v) is that (u v) arrives the distance of Fourier domain initial point to point to D; N is the exponent number of wave filter;
2) above-mentioned pretreated image is carried out binary conversion treatment;
3) window that is respectively m * 9 and 9 * n with two sizes longitudinally, the transversal scanning entire image;
4) in the window field of m * 9, calculate
X
iBe row pixel value (being 1 or 0), S in bianry image
jBe a row pixel value and, Sump
RowFor in the window each row and minimum value; Work as Sump
RowValue during less than predetermined threshold value T, judge that in the window this classify the element in the dark border line as, and these row are decided to be the central series of m * 9 windows, m * 9 window center are listed as for this, filling the interior all pixels of window is 1 (1 for white), finishes the part deletion of frame line;
5) in the window field of 9 * n, calculate
X
jBe row pixel value (being 1 or 0), S here
iFor the one-row pixels value and, Sump
OKFor in the window each the row and minimum value; Work as Sump
OKValue during less than threshold value, fill in the window all that pixels are 1 (1 for white), finish the deletion of frame line.
The size of described threshold value T is decided to be about 20% of a row pixel summation.
It is identical and less than the row or column of threshold value to occur the above minimum value of two rows in window, fills as window center near the row or column of picture centre position selecting.
Beneficial effect
The present invention proposes a kind of filtering method of fingerprint frame line.Frame line is during from printing ink stamp fingerprint, and the printing sorting track on the capture card has been brought into database in scanning process.These lines have higher density, and are connected with the fingerprint line.Utilize this method can realize framing mask is handled, to eliminate the influence of frame to the output result.
Description of drawings
Fig. 1: frame influences the center point location
A: the fingerprint central point location of no frame line; B: frame line makes the central point error in judgement
Fig. 2: the frame line in the fingerprint image
Fig. 3: vertical, horizontal scanning window
Fig. 4: frame filtration result
A: be original image; B: be the bandreject filtering bianry image
C: be deletion frame line image; D: be the correct positioning image of central point
Embodiment
Now in conjunction with the accompanying drawings the present invention is further described:
The hardware environment that is used to implement is: Pentium-2.66G computing machine, 256MB internal memory, 64M video card, the software environment of operation is: Window XP.We have realized the method that the present invention proposes with VC++ and MATLAB 7.1 programming languages.
At first image is adopted the bandreject filtering pre-service:
D in the formula
l, D
hBe cutoff frequency; (u v) is that (u v) arrives the distance of Fourier domain initial point to point to D; N is the exponent number of wave filter.D here
l, D
kChoose according to the situation of different images is adaptive and choose.According to the method in the process flow diagram image is carried out binary conversion treatment then.
Our characteristic of analysis image frame straight line at first after image is through pre-service such as filtering, two-value, be the frame horizontal line that in the image is run through image after the binaryzation as shown in Figure 4, although this line has some inclinations, edge also irregular, from character matrix we always can to find one-row pixels be that be 0 entirely (0 expression is black) or the overwhelming majority is 0 array.Simultaneously, constitute because fingerprint is spaced crestal line and valley line, and be present under the white background, thus in array laterally or longitudinally of streakline, the distribution of black picture element less than this pixelated array and 50%.These have created condition for our selected threshold realizes the deletion of frame line.Vertical line in the similar image has this characteristic too.
It for a width of cloth size 640 * 640 fingerprint image, we find through statistics, the width mean pixel of frame line is 6, because there is certain deviation in the minimum value in this line on level or the vertical direction and the axle center of this line, so it is when selecting the width of filter window, bigger slightly than streakline width.The window width that we select is 9., be the image of m * n for width of cloth size, need handle image respectively with the window that two sizes are respectively m * 9 and 9 * n, with these two windows longitudinally, horizontal scanning entire image.
In the window field of m * 9, calculate every row and:
X
iBe row pixel value (being 1 or 0), S in bianry image
jBe a row pixel value and, Sump
RowFor in the window each row and minimum value.
Work as Sump
RowValue during less than a certain predetermined threshold value T, judge that in the window this classify the element in the dark border line as, and these row are decided to be the central series of m * 9 windows, m * 9 window center are listed as for this, filling the interior all pixels of window is 1 (1 for white), thereby finishes the deletion of frame line.Threshold value T chooses analysis according to the front, and size is decided to be about 20% of a row pixel summation, that is: T=0.2m (T is an empirical value).
Equally, in the window field of 9 * n, calculate every row pixel and:
X
jBe row pixel value (being 1 or 0), S here
iFor the one-row pixels value and, Sump
OKFor in the window each the row and minimum value.
Work as Sump
OKValue during less than threshold value, fill in the window all that pixels are 1 (1 for white), finish the deletion of frame line.
It is identical and less than the row or column of threshold value that the present invention points out to occur sometimes in the window the above minimum value of two rows, at this time will select to fill as window center near the row or column of picture centre position.And from the frame line to the image border, have some streaklines or other ink marks sometimes, these still exist after the filtering straight line, streakline outside the angle frame line of dactylography is invalid in extract minutiae, therefore also should be these together filterings in the filtering straight line.The row or column at the minimum value place that calculates in window then all is filled to background from the window's position to image coboundary or left hand edge less than the picture centre position; Otherwise then all be filled to background to image lower limb or right hand edge.
Claims (3)
1. the filtering method of a fingerprint frame line is characterized in that step is as follows:
1) at first to carrying out the bandreject filtering pre-service for the image of m * n for width of cloth size:
D in the formula
l, D
hBe cutoff frequency; (u v) is that (u v) arrives the distance of Fourier domain initial point to point to D; N is the exponent number of wave filter;
2) above-mentioned pretreated image is carried out binary conversion treatment;
3) window that is respectively m * 9 and 9 * n with two sizes longitudinally, the transversal scanning entire image;
4) in the window field of m * 9, calculate
X
iBe row pixel value (being 1 or 0), S in bianry image
jBe a row pixel value and, Sump
RowFor in the window each row and minimum value; Work as Sump
RowValue during less than predetermined threshold value T, judge that in the window this classify the element in the dark border line as, and these row are decided to be the central series of m * 9 windows, m * 9 window center are listed as for this, filling the interior all pixels of window is 1 (1 for white), finishes the part deletion of frame line;
5) in the window field of 9 * n, calculate
X
jBe row pixel value (being 1 or 0), S here
iFor the one-row pixels value and, Sump
OKFor in the window each the row and minimum value; Work as Sump
OKValue during less than threshold value, fill in the window all that pixels are 1 (1 for white), finish the deletion of frame line.
2. the filtering method of fingerprint frame line according to claim 1, it is characterized in that: the size of described threshold value T is decided to be about 20% of a row pixel summation.
3. the filtering method of fingerprint frame line according to claim 1 is characterized in that: it is identical and less than the row or column of threshold value to occur the above minimum value of two rows in window, fills as window center near the row or column of picture centre position selecting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007100179280A CN100573552C (en) | 2007-05-24 | 2007-05-24 | The filtering method of fingerprint frame line |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007100179280A CN100573552C (en) | 2007-05-24 | 2007-05-24 | The filtering method of fingerprint frame line |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101286194A true CN101286194A (en) | 2008-10-15 |
| CN100573552C CN100573552C (en) | 2009-12-23 |
Family
ID=40058395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2007100179280A Expired - Fee Related CN100573552C (en) | 2007-05-24 | 2007-05-24 | The filtering method of fingerprint frame line |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100573552C (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108271061A (en) * | 2016-12-30 | 2018-07-10 | 央视国际网络无锡有限公司 | A kind of method for being inserted into high contrast frame subtitle in video |
| CN113361448A (en) * | 2021-06-23 | 2021-09-07 | 浙江中正智能科技有限公司 | Method for detecting interference straight line in fingerprint image |
-
2007
- 2007-05-24 CN CNB2007100179280A patent/CN100573552C/en not_active Expired - Fee Related
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108271061A (en) * | 2016-12-30 | 2018-07-10 | 央视国际网络无锡有限公司 | A kind of method for being inserted into high contrast frame subtitle in video |
| CN113361448A (en) * | 2021-06-23 | 2021-09-07 | 浙江中正智能科技有限公司 | Method for detecting interference straight line in fingerprint image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100573552C (en) | 2009-12-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2708422C1 (en) | Atm management system and method | |
| CN109657632B (en) | Lane line detection and identification method | |
| Raja | Fingerprint recognition using minutia score matching | |
| CN100380391C (en) | Method and apparatus for identifying a palmprint image | |
| CN103530878B (en) | A kind of edge extracting method based on convergence strategy | |
| CN108388896A (en) | A kind of licence plate recognition method based on dynamic time sequence convolutional neural networks | |
| CN109871938A (en) | A kind of components coding detection method based on convolutional neural networks | |
| CN107045634B (en) | Text positioning method based on maximum stable extremum region and stroke width | |
| CN107958201A (en) | A kind of intelligent checking system and method for vehicle annual test insurance policy form | |
| CN101604384B (en) | Individualized fingerprint identification method | |
| CN102629322B (en) | Character feature extraction method based on stroke shape of boundary point and application thereof | |
| CN102254172B (en) | Method for segmenting fingerprint image based on cellular neural network and morphology | |
| CN110766016B (en) | Code-spraying character recognition method based on probabilistic neural network | |
| CN107657209B (en) | Template image registration mechanism based on finger vein image quality | |
| CN106778668A (en) | A kind of method for detecting lane lines of the robust of joint RANSAC and CNN | |
| CN110110608B (en) | Forklift speed monitoring method and system based on vision under panoramic monitoring | |
| CN107066952A (en) | A kind of method for detecting lane lines | |
| CN111008651B (en) | Image reproduction detection method based on multi-feature fusion | |
| CN104766078A (en) | Slant angle detecting method and slant angle detecting device of target image | |
| CN105354558A (en) | Face image matching method | |
| CN107490583A (en) | A kind of intermediate plate defect inspection method based on machine vision | |
| CN108171098A (en) | A kind of code detection method and equipment | |
| CN115631128A (en) | Circular medicine quantity detection method based on circular power theorem | |
| CN100573552C (en) | The filtering method of fingerprint frame line | |
| CN106342325B (en) | A kind of region segmentation method of fingerprint image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20091223 Termination date: 20120524 |