CN107437064A - Living iris detection method based on spectrum analysis - Google Patents
Living iris detection method based on spectrum analysis Download PDFInfo
- Publication number
- CN107437064A CN107437064A CN201710540599.1A CN201710540599A CN107437064A CN 107437064 A CN107437064 A CN 107437064A CN 201710540599 A CN201710540599 A CN 201710540599A CN 107437064 A CN107437064 A CN 107437064A
- Authority
- CN
- China
- Prior art keywords
- iris
- living body
- spectrum
- image
- body iris
- 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
- G06V40/40—Spoof detection, e.g. liveness detection
- G06V40/45—Detection of the body part being alive
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V40/18—Eye characteristics, e.g. of the iris
- G06V40/193—Preprocessing; Feature extraction
Abstract
The invention discloses a kind of living iris detection method based on spectrum analysis, methods described includes living body iris spectrum distribution features training process and iris In vivo detection process.The present invention utilizes living body iris and the spectrum distribution feature difference for printing iris, train the spectrum distribution feature for the living body iris for being different from printing iris, realize living body iris with printing the classification of iris, effectively prevent the prosthese of counterfeit living body iris from attacking phenomenon in authentication procedures, suitable for safeguarding user information safety.The present invention sets a variety of iris In vivo detection decision mechanisms, can configure according to the actual requirements.
Description
Technical field
The present invention relates to the technical fields such as Digital Image Processing, pattern-recognition and statistical learning, and in particular to based on frequency spectrum
The living iris detection method of analysis.
Background technology
The identity of people is identified by the texture difference between analyzing different irises for iris recognition and certification, has only
The advantages that one property is high, stability is strong, the non-property invaded, has been successfully applied to the identity authentication of the occasions such as airport, customs, bank.
However, gradually stepping up with popularity, iris authentication system is also faced with the threat and attack of various artificial forgery technologies.
If iris authentication system can not be detected and alarmed to false iris data exactly, potential damage will be caused to authorized user
Lose.
Therefore, how fast and effeciently to realize that iris In vivo detection is still one urgently to be resolved hurrily in iris authentication system
Problem.
The content of the invention
The present invention seeks to for problems of the prior art, there is provided a kind of fast and effectively iris In vivo detection side
Method.
A kind of living iris detection method based on spectrum analysis, methods described are instructed including living body iris spectrum distribution feature
Practice process S1 and iris In vivo detection process S2.
The living body iris spectrum distribution features training process S1 includes following sub-step:Living body iris database is built to make
To train storehouse, the image in training storehouse is pre-processed, cuts out ocular;Utilize Spectrum Conversion method extraction eye area
Characteristic of field, draw eye spectrogram;The error image of the binaryzation of eye spectrogram is obtained using image processing method;Carry out area
Domain analysis draws eye spectrum analysis figure;The eye spectrum analysis figure of all images in superposition training storehouse, statistics living body iris frequency
Spectral power distribution figure;Multi-level decomposition is carried out to living body iris spectrum energy distribution map, draws multistage living body iris spectrum energy point
Butut.
The iris In vivo detection process S2 includes following sub-step:Test image is pre-processed, cuts out eye
Image;Ocular feature is extracted using Spectrum Conversion method, draws eye spectrogram;Eye is obtained using image processing method
The error image of the binaryzation of spectrogram;According to the multistage living body iris spectrum energy distribution map of training gained, judgement machine is chosen
System carries out In vivo detection;Export In vivo detection result.
Preferably, the living body iris spectrum distribution features training process S1 is comprised the following specific steps that:
S11:Image in training storehouse is pre-processed, eye position is positioned using eye detection method, cuts out eye
Portion's image;
S12:The spectrum signature of ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), obtains eye frequency
Spectrogram;
S13:Image procossing, including filtering, difference processing and binary conversion treatment are carried out to eye spectrogram, obtains eye frequency
The error image of the binaryzation of spectrogram;
S14:Regional analysis is carried out to the error image of binaryzation using connected domain analysis method, retain connected domain area compared with
Big region, obtain spectrum analysis figure;
S15:The eye spectrum analysis figure of all images, counts living body iris spectrum energy distribution map in superposition training storehouse;
S16:Multi-level decomposition is carried out to living body iris spectrum energy distribution map, draws strong, weak living body iris spectrum energy point
Butut.
Wherein, step S16 multi-level decomposition method is:Threshold value a and b, wherein a > b, by living body iris spectrum energy are set
Zone marker of the distribution map numerical value more than b is 1, and otherwise labeled as 0, this mark image is designated as into weak living body iris spectrum energy point
Butut;It is 1 by zone marker of the living body iris spectrum energy distribution map numerical value more than a, otherwise labeled as 0, by this mark image
It is designated as strong living body iris spectrum energy distribution map.
Preferably, the iris In vivo detection process S2 is comprised the following specific steps that:
S21:Test image is pre-processed, eye position is positioned using eye detection method, cuts out eyes image;
S22:The spectrum signature of ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), obtains eye frequency
Spectrogram;
S23:Image procossing, including filtering, difference processing and binary conversion treatment are carried out to eye spectrogram, obtains eye frequency
The error image of the binaryzation of spectrogram;
S24:According to the multistage living body iris spectrum energy distribution map of training gained, choose decision mechanism and carry out live body inspection
Survey;
S25:Export In vivo detection result.
Preferably, step S24 decision mechanism is arranged to following steps in the iris In vivo detection process S2:
S2411:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step
The error image of binaryzation is filtered processing obtained by S23, and carries out connected domain analysis, retains the larger area of connected domain area
Domain;
S2412:Weak living body iris is determined whether, if it is, continuing strong living body iris detection, into step
S2413, otherwise it is judged to detect prosthese, sends alarm, iris In vivo detection process terminates;
S2413:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step
Image after S2411 filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2414:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, iris is lived
Body detection process terminates, and otherwise return to step S21 carries out iris In vivo detection to next width test image.
Preferably, step S24 decision mechanism may be arranged as following steps in the iris In vivo detection process S2:
S2421:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step
The error image of binaryzation is filtered processing obtained by S23, and carries out connected domain analysis, retains the larger area of connected domain area
Domain;
S2422:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, iris is lived
Body detection process terminates, and otherwise continues weak living body iris detection, into step S2423;
S2423:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step
Image after S2421 filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2424:Determine whether weak living body iris, if it is, carrying out iris In vivo detection to next two field picture, return
Step S21 carries out iris In vivo detection to next width test image, is otherwise judged to detect prosthese, sends alarm, iris live body
Detection process terminates.
Preferably, the step S2422 of decision mechanism is being judged to after not being living body iris, accumulative continuously to be judged to not be live body
The number of iris, when the number exceedes predetermined threshold, prosthese is directly judged to, sends alarm, iris In vivo detection process terminates.
Preferably, step S13 and step S23 filtering process include filtering twice, use radius as 1 pixel
Wave filter is once filtered to image, uses radius to carry out secondary filtering to image for the wave filter of 24 pixels.
Preferably, step S13 and step S23 difference processing method be to once filtering, secondary filtering image pixel by pixel
Make the difference, take absolute value, multiplied by with zoom factor.
The beneficial effects of the invention are as follows:The present invention utilizes living body iris and the spectrum distribution feature difference for printing iris, instruction
The spectrum distribution feature for the living body iris for being different from printing iris is practised, living body iris and printing iris classification are realized, in identity
The prosthese attack phenomenon of counterfeit living body iris is detected in verification process, suitable for safeguarding user information safety.The present invention is set
A variety of iris In vivo detection decision mechanisms, can be configured according to the actual requirements.
It should be appreciated that foregoing description substantially and follow-up description in detail are exemplary illustration and explanation, should not
As the limitation to the claimed content of the present invention.
Brief description of the drawings
With reference to the accompanying drawing enclosed, the present invention more purpose, function and advantages will pass through the as follows of embodiment of the present invention
Description is illustrated, wherein:
Fig. 1 shows a kind of living iris detection method flow chart based on spectrum analysis of the present invention;
Fig. 2 a show a kind of iris In vivo detection first embodiment iris In vivo detection side based on spectrum analysis of the present invention
Method;
Fig. 2 b show a kind of iris In vivo detection second embodiment iris In vivo detection side based on spectrum analysis of the present invention
Method;
Fig. 2 c show a kind of iris In vivo detection 3rd embodiment iris In vivo detection side based on spectrum analysis of the present invention
Method;
Fig. 3 a schematically show living body iris according to iris In vivo detection of the present invention;
Fig. 3 b schematically show printing iris according to iris In vivo detection of the present invention;
Fig. 3 c are spectrogram corresponding to the living body iris 3a according to iris In vivo detection of the present invention;
Fig. 3 d are spectrogram corresponding to the printing iris 3b according to iris In vivo detection of the present invention.
Embodiment
By reference to one exemplary embodiment, the purpose of the present invention and function and the side for realizing these purposes and function
Method will be illustrated.However, the present invention is not limited to one exemplary embodiment as disclosed below;Can by multi-form come
It is realized.The essence of specification is only to aid in the detail of the various equivalent modifications Integrated Understanding present invention.
Hereinafter, embodiments of the invention will be described with reference to the drawings.In the accompanying drawings, identical reference represents identical
Or similar part, or same or like step.
Fig. 1 is the basic step of the living iris detection method based on spectrum analysis, as shown in figure 1, methods described includes
The living body iris spectrum distribution features training process S1 and iris In vivo detection process S2 based on spectrum analysis, the living body iris
Spectrum distribution features training process S1 is comprised the following specific steps that:
Step S11 is that the image in training storehouse is pre-processed, and positions eye position using eye detection method, cuts
Go out ocular.
Step S12 is the spectrum signature that ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), is obtained
Eye spectrogram, and image enhancement processing is further done, image enhaucament threshold value T1 is set, default threshold will be less than in eye spectrogram
The pixel value of the point of value is arranged to T1.
Step S13 is that image procossing is carried out to eye spectrogram, including filtering, difference processing and binary conversion treatment, is obtained
The error image of the binaryzation of eye spectrogram.According to the symmetry of spectrogram, only the upper half area of spectrogram can be carried out
Filtering process.Filter processing method includes mean filter and medium filtering etc., has filtering twice, uses radius as 1
The wave filter of pixel is once filtered to image, uses radius to carry out secondary filtering to image for the wave filter of 24 pixels.
Difference processing method be to once filtering, secondary filtering image pixel by pixel make the difference, take absolute value, multiplied by with zoom factor.
Step S14 is to carry out regional analysis to the error image of binaryzation using connected domain analysis method, retains connected domain
The larger region of area, obtains spectrum analysis figure.
Step S15 is the eye spectrum analysis figure of all images in superposition training storehouse, and statistics living body iris spectrum energy divides
Butut.Living body iris spectrum energy distribution map represents the frequency that living body iris spectrum energy occurs, living body iris spectrum energy point
The numerical value of a certain position is bigger in Butut, represent the position occur living body iris spectrum energy probability it is higher.It is that is, living
The spectrum energy of body iris is concentrated in the higher regional extent of living body iris spectrum energy distribution map numerical value;Print the frequency of iris
Spectrum energy prints the spectrum energy point of iris not in the higher regional extent of living body iris spectrum energy distribution map numerical value
The energy area of multiple aggregations in Butut be present.
Step S16 is to carry out multi-level decomposition to living body iris spectrum energy distribution map, is arranged to two-stage, draws strong, weak work
Body iris spectrum energy distribution map.Specific method is to set threshold value a and b, wherein a > b, by living body iris spectrum energy distribution map
Zone marker of the numerical value more than b is 1, and otherwise labeled as 0, this mark image is designated as into weak living body iris spectrum energy distribution map;
It is 1 by zone marker of the living body iris spectrum energy distribution map numerical value more than a, otherwise labeled as 0, this mark image is designated as by force
Living body iris spectrum energy distribution map.
The region representation living body iris and part printing iris that numerical value is 1 in weak living body iris spectrum energy distribution map
The position that spectrum energy occurs, numerical value is that 0 region representation prints the frequency spectrum of iris in weak living body iris spectrum energy distribution map
The position that energy occurs.
The position that the spectrum energy for the region representation living body iris that numerical value is 1 occurs in strong living body iris spectrum energy distribution map
Put, numerical value is that 0 region representation prints the frequency spectrum of iris and part living body iris in strong living body iris spectrum energy distribution map
The position that energy occurs.
The iris In vivo detection process S2 based on spectrum analysis is comprised the following specific steps that:
Step S21 positions eye position to be pre-processed to test image, using eye detection method, cuts out eye
Image.If being not detected by eyes, next width test image can be pre-processed.
Step S22 is the spectrum signature that ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), is obtained
Eye spectrogram, and image enhancement processing is further done, image enhaucament threshold value T1 is set, default threshold will be less than in eye spectrogram
The pixel value of the point of value is arranged to T1.
Step S23 is that image procossing is carried out to eye spectrogram, including filtering, difference processing and binary conversion treatment, is obtained
The error image of the binaryzation of eye spectrogram.According to the symmetry of spectrogram, only the upper half area of spectrogram can be carried out
Filtering process.Filter processing method includes mean filter and medium filtering etc., has filtering twice, uses radius as 1
The wave filter of pixel is once filtered to image, uses radius to carry out secondary filtering to image for the wave filter of 24 pixels.
Difference processing method be to once filtering, secondary filtering image pixel by pixel make the difference, take absolute value, multiplied by with zoom factor.
Step S24 is according to the multistage living body iris spectrum energy distribution map of training gained, chooses decision mechanism and is lived
Physical examination is surveyed, and the decision mechanism includes 3 kinds.
Step S25 is output In vivo detection result, in iris recognition identification verification process, if output is live body rainbow
Film, then active user can continue iris registration and iris recognition, if output is prosthese, have no right to carry out iris registration
And iris recognition.
Embodiment 1
1st kind of decision mechanism passes through first order inspection policies for there is prosthese attack than more frequently application scenarios
The detection of weak living body iris detects whether as prosthese first, the situation of a large amount of prostheses attacks can be so excluded, afterwards for few
Several irises for being judged to weak live body carry out the strong living body iris detection of second level inspection policies again, finally detect whether as live body rainbow
Film.The beneficial effect of this inspection policies is the situation that prosthese attack is only can detect that by first order inspection policies,
In authentication procedures, can be rapidly and efficiently alarm is sent to prosthese attack, ensure the security of authentication.Specifically
Step such as Fig. 2 a S2411~S2414:
S2411:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step
The error image of binaryzation is filtered processing obtained by S23, and carries out connected domain analysis, retains the larger area of connected domain area
Domain;
S2412:Weak living body iris is determined whether, if it is, continuing strong living body iris detection, into step
S2413, otherwise it is judged to detect prosthese, sends alarm, iris In vivo detection process terminates;
S2413:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step
Image after S2411 filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2414:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, iris is lived
Body detection process terminates, and otherwise return to step S21 carries out iris In vivo detection to next width test image.
Embodiment 2
2nd kind of decision mechanism is directed to the application scenarios for prosthese attack seldom occur, passes through the strong work of first order inspection policies
Body iris detection detects whether that, for living body iris, the image that remaining small part can not be judged to living body iris carries out second again first
The weak living body iris detection of level inspection policies, is finally detected whether as prosthese.The beneficial effect of this inspection policies is only
Living body iris is can detect that by first order inspection policies, the corresponding authority of user is assigned and carries out authentication, for live body
Iris authenticating user identification process reduces the computation complexity of algorithm.Specific steps such as Fig. 2 b S2421~S2424:
S2421:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step
The error image of binaryzation is filtered processing obtained by S23, and carries out connected domain analysis, retains the larger area of connected domain area
Domain;
S2422:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, iris is lived
Body detection process terminates, and otherwise continues weak living body iris detection, into step S2423;
S2423:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step
Image after S2421 filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2424:Determine whether weak living body iris, if it is, carrying out iris In vivo detection to next two field picture, return
Step S21 carries out iris In vivo detection to next width test image, is otherwise judged to detect prosthese, sends alarm, iris live body
Detection process terminates.
Embodiment 3
3rd kind of decision mechanism is compared with the 2nd kind of decision mechanism, and the judgement for living body iris is more harsh, i.e., first
After the strong living body iris detection of level inspection policies, if continuous be judged as it not being that the number of strong living body iris exceedes default threshold
Value, then directly it is judged to print iris, and alert, In vivo detection no longer is carried out to next two field picture of active user.
Specific steps such as Fig. 2 c S2431~S2436:
S2431:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step
The error image of binaryzation is filtered processing obtained by S23, and carries out connected domain analysis, retains the larger area of connected domain area
Domain;
S2432:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, iris is lived
Body detection process terminates, otherwise into step S2433;
S2433:It is accumulative continuous be judged as be not living body iris number;
S2434:Judge it is continuous be judged as it not being whether the number of living body iris exceedes threshold value, if it exceeds being then judged to detect
To prosthese, alarm is sent, iris In vivo detection process terminates, and otherwise continues weak living body iris detection, into step
S2435;
S2435:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step
Image after S2431 filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2436:Determine whether weak living body iris, if it is, carrying out iris In vivo detection to next two field picture, return
Step S21 carries out iris In vivo detection to next width test image, is otherwise judged to detect prosthese, sends alarm, iris live body
Detection process terminates.
Fig. 3 a~Fig. 3 d are frequency spectrum illustrated example of the living body iris with printing iris, wherein, Fig. 3 a are living body iris, and Fig. 3 b are
Printing iris from living body iris shown in Fig. 3 a, Fig. 3 c, Fig. 3 d are respectively Fig. 3 a, Fig. 3 b Fast Fourier Transform (FFT) frequency spectrum
Figure, it can be seen that the spectrum distribution of living body iris and printing iris has very big difference:High frequency in the spectrogram of living body iris
Component distribution is more concentrated, and compared with the frequency spectrum of living body iris, printing iris has more medium, high frequency frequency components, Fig. 3 d
In regular four bright spots for being distributed in frequency spectrum surrounding be frequecy characteristic caused by print procedure.Therefore according to frequency spectrum point
Cloth difference can distinguish living body iris and printing iris, the position of living body iris frequency spectrum appearance be filtered out first, afterwards according to surplus
The number in complementary energy region is judged, sets threshold value, if the number in dump energy region is higher than threshold value, is judged to print rainbow
Film, otherwise it is judged to living body iris.
With reference to the explanation of the invention disclosed here and practice, other embodiment of the invention is for those skilled in the art
It all will be readily apparent and understand.Illustrate and embodiment is to be considered only as exemplary, of the invention true scope and purport is equal
It is defined in the claims.
Claims (9)
1. a kind of living iris detection method based on spectrum analysis, methods described includes living body iris spectrum distribution features training
Process S1 and iris In vivo detection process S2;
Wherein, living body iris spectrum distribution features training process S1 includes following sub-step:Build living body iris database conduct
Storehouse is trained, the image in training storehouse is pre-processed, cuts out ocular;Ocular is extracted using Spectrum Conversion method
Feature, draw eye spectrogram;The error image of the binaryzation of eye spectrogram is obtained using image processing method;Carry out region
Analysis draws eye spectrum analysis figure;The eye spectrum analysis figure of all images, counts living body iris frequency spectrum in superposition training storehouse
Energy profile;Multi-level decomposition is carried out to living body iris spectrum energy distribution map, draws multistage living body iris spectrum energy distribution
Figure;
The iris In vivo detection process S2 includes following sub-step:Test image is pre-processed, cuts out eyes image;
Ocular feature is extracted using Spectrum Conversion method, draws eye spectrogram;Eye frequency spectrum is obtained using image processing method
The error image of the binaryzation of figure;According to the multistage living body iris spectrum energy distribution map of training gained, choose decision mechanism and enter
Row In vivo detection;Export In vivo detection result.
A kind of 2. living iris detection method based on spectrum analysis according to claim 1, it is characterised in that the work
Body iris spectrum distribution features training process S1 is comprised the following specific steps that:
S11:Image in training storehouse is pre-processed, eye position is positioned using eye detection method, cuts out eye figure
Picture;
S12:The spectrum signature of ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), obtains eye frequency spectrum
Figure;
S13:Image procossing, including filtering, difference processing and binary conversion treatment are carried out to eye spectrogram, obtains eye spectrogram
Binaryzation error image;
S14:Regional analysis is carried out to the error image of binaryzation using connected domain analysis method, it is larger to retain connected domain area
Region, obtain spectrum analysis figure;
S15:The eye spectrum analysis figure of all images, counts living body iris spectrum energy distribution map in superposition training storehouse;
S16:Multi-level decomposition is carried out to living body iris spectrum energy distribution map, draws strong, weak living body iris spectrum energy distribution map.
3. step S16 multi-level decomposition side in living body iris spectrum distribution features training process S1 according to claim 2
Method is:Threshold value a and b are set, they wherein a > b, are 1 by zone marker of the living body iris spectrum energy distribution map numerical value more than b, and it is no
0 is then labeled as, this mark image is designated as weak living body iris spectrum energy distribution map;By living body iris spectrum energy distribution map number
Zone marker of the value more than a is 1, and otherwise labeled as 0, this mark image is designated as into strong living body iris spectrum energy distribution map.
A kind of 4. living iris detection method based on spectrum analysis according to claim 1, it is characterised in that the iris
In vivo detection process S2 is comprised the following specific steps that:
S21:Test image is pre-processed, eye position is positioned using eye detection method, cuts out eyes image;
S22:The spectrum signature of ocular is extracted using the Spectrum Conversion method of Fast Fourier Transform (FFT), obtains eye frequency spectrum
Figure;
S23:Image procossing, including filtering, difference processing and binary conversion treatment are carried out to eye spectrogram, obtains eye spectrogram
Binaryzation error image;
S24:According to the multistage living body iris spectrum energy distribution map of training gained, choose decision mechanism and carry out In vivo detection;
S25:Export In vivo detection result.
5. the step S24 in iris In vivo detection process S2 according to claim 4, it is characterised in that the judgement machine
Set up and be set to following steps:
S2411:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to iris live body
The error image of binaryzation is filtered processing obtained by step S23 in detection process S2, and carries out connected domain analysis, retains connection
The larger region of domain area;
S2412:Determine whether weak living body iris, if it is, continue the detection of strong living body iris, into step S2413,
Otherwise it is judged to detect prosthese, sends alarm, iris In vivo detection process terminates;
S2413:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to step S2411
Image after filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2414:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, the inspection of iris live body
Survey process terminates, and the step S21 otherwise returned in claim 4 carries out iris In vivo detection to next width test image.
6. step S24 in iris In vivo detection process S2 according to claim 4, it is characterised in that the decision mechanism
It may be arranged as following steps:
S2421:Strong living body iris detection is carried out, using the strong living body iris spectrum energy distribution map of training gained, to iris live body
The error image of binaryzation is filtered processing obtained by step S23 in detection process S2, and carries out connected domain analysis, retains connection
The larger region of domain area;
S2422:Determine whether strong living body iris, if it is, being judged to living body iris, export testing result, the inspection of iris live body
Survey process terminates, and otherwise continues weak living body iris detection, into step S2423;
S2423:Weak living body iris detection is carried out, using the weak living body iris spectrum energy distribution map of training gained, to step S2421
Image after filtering process is filtered processing, and carries out connected domain analysis, retains the larger region of connected domain area;
S2424:Determine whether weak living body iris, if it is, carrying out iris In vivo detection to next two field picture, return to iris
Step S21 in In vivo detection process S2 carries out living body iris detection to next width test image, is otherwise judged to detect prosthese,
Alarm is sent, iris In vivo detection process terminates.
7. iris In vivo detection process S2 according to claim 6, it is characterised in that the step of the decision mechanism
S2422 is being judged to after not being living body iris, it is accumulative it is continuous be judged to be not living body iris number, when the number exceedes predetermined threshold
During value, prosthese is directly judged to, sends alarm, iris In vivo detection process terminates.
A kind of 8. living iris detection method based on spectrum analysis according to claim 2 or 4, it is characterised in that institute
State the filtering of step S23 in step S13 and iris In vivo detection process S2 in living body iris spectrum distribution features training process S1
Processing includes filtering twice, uses radius once to be filtered to image for the wave filter of 1 pixel, uses radius as 24
The wave filter of individual pixel carries out secondary filtering to image.
A kind of 9. living iris detection method based on spectrum analysis according to claim 2 or 4, it is characterised in that institute
State the difference of step S23 in step S13 and iris In vivo detection process S2 in living body iris spectrum distribution features training process S1
Processing method is that the once filtering described in the living iris detection method based on spectrum analysis, secondary filtering image pixel by pixel are done
Difference, take absolute value, multiplied by with zoom factor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710540599.1A CN107437064B (en) | 2017-07-05 | 2017-07-05 | Iris living body detection method based on spectral analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710540599.1A CN107437064B (en) | 2017-07-05 | 2017-07-05 | Iris living body detection method based on spectral analysis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107437064A true CN107437064A (en) | 2017-12-05 |
CN107437064B CN107437064B (en) | 2020-11-20 |
Family
ID=60460852
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710540599.1A Active CN107437064B (en) | 2017-07-05 | 2017-07-05 | Iris living body detection method based on spectral analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107437064B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109472768A (en) * | 2018-09-19 | 2019-03-15 | 上海泛洲信息科技有限公司 | A method of in kind and non-flat image in kind is distinguished using spectrum analysis |
CN109684966A (en) * | 2018-12-17 | 2019-04-26 | 北京中科虹霸科技有限公司 | A kind of iris capturing apparatus and method for of extended field of view range |
CN111507198A (en) * | 2020-03-25 | 2020-08-07 | 北京万里红科技股份有限公司 | Training method for printed iris detection model, and printed iris detection method and device |
WO2022156441A1 (en) * | 2021-01-21 | 2022-07-28 | 展讯通信(上海)有限公司 | Living body detection method and apparatus, storage medium, and terminal |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1426760A (en) * | 2001-12-18 | 2003-07-02 | 中国科学院自动化研究所 | Identity discriminating method based on living body iris |
-
2017
- 2017-07-05 CN CN201710540599.1A patent/CN107437064B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1426760A (en) * | 2001-12-18 | 2003-07-02 | 中国科学院自动化研究所 | Identity discriminating method based on living body iris |
Non-Patent Citations (7)
Title |
---|
LI: "Live face detection based on the analysis of Fourier spectra", 《PROCEEDINGS OF SPIE》 * |
RUI CHEN: "Liveness detection for iris recognition using multispectral images", 《PATTERN RECOGNITION LETTERS》 * |
XIAOFU HE: "A Fake Iris Detection Method Based on FFT and Quality Assessment", 《2008 CHINESE CONFERENCE ON PATTERN RECOGNITION》 * |
余龙华: "人眼检测及瞳孔定位", 《计算机工程与应用》 * |
吴国龙: "非理想成像条件下的虹膜定位方法", 《计算机工程与设计》 * |
周水子: "虹膜识别算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
姚翠莉: "虹膜图像质量评估算法的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109472768A (en) * | 2018-09-19 | 2019-03-15 | 上海泛洲信息科技有限公司 | A method of in kind and non-flat image in kind is distinguished using spectrum analysis |
CN109472768B (en) * | 2018-09-19 | 2022-02-25 | 上海泛洲信息科技有限公司 | Method for distinguishing object and non-object plane images by using frequency spectrum analysis |
CN109684966A (en) * | 2018-12-17 | 2019-04-26 | 北京中科虹霸科技有限公司 | A kind of iris capturing apparatus and method for of extended field of view range |
CN111507198A (en) * | 2020-03-25 | 2020-08-07 | 北京万里红科技股份有限公司 | Training method for printed iris detection model, and printed iris detection method and device |
CN111507198B (en) * | 2020-03-25 | 2023-11-28 | 北京万里红科技有限公司 | Training method for printing iris detection model, and printing iris detection method and device |
WO2022156441A1 (en) * | 2021-01-21 | 2022-07-28 | 展讯通信(上海)有限公司 | Living body detection method and apparatus, storage medium, and terminal |
Also Published As
Publication number | Publication date |
---|---|
CN107437064B (en) | 2020-11-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pacut et al. | Aliveness detection for iris biometrics | |
CN107437064A (en) | Living iris detection method based on spectrum analysis | |
CN101923640B (en) | Method for distinguishing false iris images based on robust texture features and machine learning | |
Czajka | Database of iris printouts and its application: Development of liveness detection method for iris recognition | |
Oh et al. | Extracting sclera features for cancelable identity verification | |
KR101443139B1 (en) | Single image-based fake face detection | |
Devi et al. | An effective texture feature extraction approach for iris recognition system | |
Rana et al. | Iris recognition system using PCA based on DWT | |
Ferrara et al. | On the impact of alterations on face photo recognition accuracy | |
KR20180003475A (en) | Method of detecting fraud of an iris recognition system | |
Bodade et al. | Fake iris detection: A holistic approach | |
CN107527423B (en) | Paper money anti-counterfeiting identification method and paper money identification device | |
Venkataramani et al. | Fingerprint verification using correlation filters | |
Shen et al. | Iris matching by crypts and anti-crypts | |
Paderes | A comparative review of biometric security systems | |
Chen et al. | Iris recognition using 3D co-occurrence matrix | |
Sharma | Fingerprint biometric system: a survey | |
Satapathy et al. | Generalized benford’s law for fake fingerprint detection | |
Seif et al. | Iris identification based on log Gabor filtering | |
Hegde et al. | Human authentication using finger knuckle print | |
Huang et al. | Iris recognition using Fourier-wavelet features | |
Swati et al. | Iris recognition using Gabor | |
Thukral et al. | IRIS spoofing through print attack using SVM classification with gabor and HOG features | |
KR100359920B1 (en) | Discrimination method for imitative iris in iris recognition system | |
DE102017126207A1 (en) | Sensor system for testing hand vein patterns |
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 |