CN205318544U - Device and system are prevented cheaing by ATM based on three dimensional face identification - Google Patents
Device and system are prevented cheaing by ATM based on three dimensional face identification Download PDFInfo
- Publication number
- CN205318544U CN205318544U CN201521126045.XU CN201521126045U CN205318544U CN 205318544 U CN205318544 U CN 205318544U CN 201521126045 U CN201521126045 U CN 201521126045U CN 205318544 U CN205318544 U CN 205318544U
- Authority
- CN
- China
- Prior art keywords
- atm
- infrared
- face
- facial image
- visible light
- 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.)
- Active
Links
- 230000001815 facial effect Effects 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims abstract description 16
- 239000000758 substrate Substances 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 6
- 238000001514 detection method Methods 0.000 abstract 2
- 238000000034 method Methods 0.000 description 15
- 238000012549 training Methods 0.000 description 8
- 230000004927 fusion Effects 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 238000000513 principal component analysis Methods 0.000 description 5
- 230000007935 neutral effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000002377 Fourier profilometry Methods 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000007634 remodeling Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Landscapes
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The utility model discloses a device and system are prevented cheaing by ATM based on three dimensional face identification, the device includes: the face detection unit for gather user's visible light facial image, three -dimensional degree of depth facial image and infrared facial image, special image processing unit is used for receiving face detection unit output user's visible light facial image, three -dimensional degree of depth facial image and infrared facial image draw corresponding diagnostic information, will diagnostic information sends backend server to go on comparing with the diagnostic information who corresponds the user who prestores to receive the comparison result that backend server returned. The utility model discloses the three -dimensional range image of end user's face, three kinds of information combination modes of people's face visible light image and people's face infrared image are compared and are judged ATM operating personnel, have promoted the operational safety of ATM.
Description
Technical field
This utility model relates to recognition of face security fields, particularly to a kind of ATM antifraud Apparatus and system based on three-dimensional face identification.
Background technology
In information technology, so-called " certification ", refer to by certain verification technique, confirm system user identity and the whole process of the digitized code name verity of system hardware (such as computer). Wherein the verification technique process of system user is called " authentication ". For petty cash is concluded the business, it is operated cost to bank counter by bank clerk too high, somewhat expensive, also inconvenient, and ATM is many due to its quantity, withdraws the money shorter without queuing or queuing time, convenient and no time limit. But because ATM is in open unserviced state, if lawless person has stolen validated user accounts information, it is possible to steal clients fund, so the safety issue being traded on ATM is very prominent.
In the use procedure of current ATM, user is first to insert in ATM by bank card, then according to the screen on ATM proposes input password, during input password, must flow through the keyboard on ATM or the PIN PAD on touch screen inputs, after password authentification is passed through, just can carry out the operation of financial business such as withdrawing deposit and transfer accounts. So lawless person wants personation validated user and carries out withdrawing the money business, it is first necessary to know the bank card information of validated user and the password of bank card, all information just having lawless person to install card reader acquisition bank card on ATM or in gate control system, then pass through duplication bank card, and by installing photographic head or the password of password recorder acquisition bank card, just can palm off validated user to be stolen away by money easily, bring property loss to the user of this card.
What common ATM adopted is all the identification authentication mode of " user name+password ", and many crimes are all by breaking through cryptoguard (usurping password, dictionary attack etc.), obtains and ATM is accessed and uses, reaches after amendment authority.Simple password authentication mode can not take precautions against external attack, can not accurately identify the true identity of offender, it is impossible to meets the safety precaution requirement of ATM system.
Owing to the confidentiality of bank card information and personal identification number is more weak, withdraw the money without card if adopted, as long as after disabled user gets the related data about card number and password, to be easy to steal deposit, it is thus desirable to a kind of technology that accounts information, password and true man's information are bundled, solve the problem that password is illegally intercepted and captured, improve the safety of ATM system.
At present, living things feature recognition, as the schemes such as fingerprint recognition, two-dimension human face identification have been used in system on ATM. But common fingerprint recognition is easily forged fingerprint film and attacks, and two-dimension human face identification is easily subject to the interference of environment light, also attacking easily by methods such as photographs, the safety of ATM system still needs to improve.
To sum up, needing one badly and prevent forgery attack, safety is high, and the simple ATM of user operation withdraws the money authenticating user identification technology.
Utility model content
The purpose of this utility model is in that to overcome above-mentioned deficiency existing in prior art, it is provided that one prevents forgery attack, and safety is high, and user operation is simply based on the ATM antifraud system of three-dimensional face identification.
In order to realize above-mentioned utility model purpose, the technical solution adopted in the utility model is:
A kind of ATM antifraud device based on three-dimensional face identification, including:
Face datection unit, for gathering the visible ray facial image of user, three dimensional depth facial image and infrared face image; Wherein visible images is gathered by visible light sensor, and infrared face image is gathered by infrared light transducer[sensor, and depth image, by the data after two infrared sensors collections, is generated by the related algorithm of patent applicant.
Processing unit, for receiving the visible ray facial image of described user of described Face datection unit unit output, three dimensional depth facial image and infrared face image, extract characteristic of correspondence information, described characteristic information is sent to background server carry out and the characteristic information comparison of the corresponding user prestored, and receives the comparison result that background server returns.
Described Face datection unit includes visible light sensor, infrared sensor and infrared projection unit;
Described visible light sensor, for gathering the visible ray facial image of described user;
Described infrared sensor, for gathering the infrared face image of described user, three dimensional depth image is the three dimensional depth facial image of user as described in respective algorithms (method our company's application number can be adopted to provide for the patent that 2015100836342 or application number are 2015100808605) gathers.
Described Face datection unit includes a substrate, and described visible light sensor, infrared sensor and infrared projection unit are installed on the substrate.
Described visible light sensor has 2, and 2 described visible light sensors are symmetrically positioned in two ends, described substrate surface left and right, and described infrared sensor and infrared projection unit are between 2 described visible light sensors.
Also include:
Display unit, for when described processing unit receives the comparison result that described background server returns, by the display output of described comparison result.
A kind of ATM antifraud system based on three-dimensional face identification, including server and the ATM antifraud device based on three-dimensional face identification as above;
The described ATM antifraud device based on three-dimensional face identification sends described characteristic information to described server;
Described server is by the characteristic information comparison of described characteristic information with the corresponding user prestored, and returns comparison result to the described ATM antifraud device based on three-dimensional face identification.
Described server, by the characteristic information comparison of described characteristic information with the corresponding user prestored, specifically includes:
Thermal infrared facial image is carried out dimension-reduction treatment, extracts the characteristic vector of thermal infrared facial image, then by the characteristic vector training by BP neutral net and identification.
The method taking Decision fusion, is weighted by summation and maximum combination, the recognition of face information obtained by PCA and LDA recognition result and confidence level is carried out Decision fusion, provides final recognition result.
Compared with prior art, the beneficial effects of the utility model:
This utility model coordinates ATM that it can be made to obtain higher safety and reduce ambient light interference. This utility model has safety height, simple and convenient, applied widely, it is possible to resolve ATM is in running, and client falls the problem that wealth that card, stolen card or the card that is cracked cause is stolen. This utility model uses face depth map, face visible images and three kinds of information of face infrared image to combine. Face thermal infrared images is not substantially affected by the impact of surrounding illumination, and depth information is possible to prevent the attack of non-genuine operator's information such as facial image. Without memory when user uses, more will not be borrowed, usurp and lose, therefore can promote the processing safety of ATM.
Accompanying drawing illustrates:
Fig. 1 is the structural representation in this utility model embodiment.
Fig. 2 is antifraud device position view on ATM that this utility model provides.
Fig. 3 is that visible images obtains schematic diagram.
Fig. 4 is that infrared light depth image obtains schematic diagram.
Fig. 5 is fusion decision model in this utility model.
Detailed description of the invention
Below in conjunction with detailed description of the invention, this utility model is described in further detail. But this should not being interpreted as, the scope of the above-mentioned theme of this utility model is only limitted to below example, and all technology realized based on this utility model content belong to scope of the present utility model.
A kind of ATM antifraud device based on three-dimensional face identification, including:
Face datection unit 1, for gathering the visible ray facial image of user, three dimensional depth facial image and infrared face image;
Processing unit 2, for receiving the visible ray facial image of described user of described Face datection unit 1 unit output, three dimensional depth facial image and infrared face image, extract characteristic of correspondence information, described characteristic information is sent to background server 4 carry out and the characteristic information comparison of the corresponding user prestored, and receives the comparison result that background server 4 returns.
Described Face datection unit 1 includes visible light sensor 11, infrared sensor 12 and infrared projection unit 13;
Described visible light sensor 11, for gathering the visible ray facial image of described user;
Described infrared sensor 12, for gathering the infrared face image of described user, meanwhile, for gathering the three dimensional depth facial image of described user when described infrared projection unit 13 projects to described user face.
Described Face datection unit 1 includes a substrate, and described visible light sensor 11, infrared sensor 12 and infrared projection unit 13 are installed on the substrate.
Described visible light sensor 11 has 2, and 2 described visible light sensors 11 are symmetrically positioned in two ends, described substrate surface left and right, described infrared sensor 12 and infrared projection unit 13 between 2 described visible light sensors 11.
Also include:
Display unit 3, for when described processing unit 2 receives the comparison result that described background server 4 returns, by the display output of described comparison result.
A kind of ATM antifraud system based on three-dimensional face identification, including server and the ATM antifraud device based on three-dimensional face identification as above;
The described ATM antifraud device based on three-dimensional face identification sends described characteristic information to described server
Described server is by the characteristic information comparison of described characteristic information with the corresponding user prestored, and returns comparison result to the described ATM antifraud device based on three-dimensional face identification.
Described server, by the characteristic information comparison of described characteristic information with the corresponding user prestored, specifically includes:
Thermal infrared facial image is carried out dimension-reduction treatment, extracts the characteristic vector of thermal infrared facial image, then by the characteristic vector training by BP neutral net and identification.
The method taking Decision fusion, is weighted by summation and maximum combination, the recognition of face information obtained by PCA and LDA recognition result and confidence level is carried out Decision fusion, provides final recognition result.
This utility model can coordinate ATM, it is prevented that forgery attack, authentication operation personnel identity, reduces the interference to face identification system of the ambient visible light line.
System is by Face datection unit 1, display unit 3, and processing unit 2 is constituted. Face datection unit 1 includes visible light sensor 11, infrared sensor 12, infrared projection unit 13, and data base on background server 4 and three-dimensional alignment algorithm.
The face authentication system of this patent utility model is embedded in ATM by the method, wherein, Face datection unit 1 is embedded in (or other can photograph the position of face) above ATM screen, display unit 3 uses ATM screen (can also additionally configure a little screen, be used for pointing out user operation and showing user's face image) and a special graphics processing unit 2(such as Fig. 2). The face information collected (is included face depth map, visible ray facial image and infrared face image three category information), 3-D view processing unit 2 processes, information (log-on message includes but not limited to the three-dimensional face features of user, 2 d texture feature, number of registration etc.) when registering with user on background server 4 after the feature obtained compares, and finally provides the decision message whether certification is passed through. Certification is not passed through, system will record guest operation and the portrait that above shoots, and refuse user and continue to operate.
3-D view processing unit 2 obtains the parallax of image according to two infrared cameras, calculates the depth image of face, fusion method is mentioned below and performs at server end.
This utility model uses face depth map, face visible images and three kinds of information of face infrared image to combine. Face thermal infrared images is not substantially affected by the impact of surrounding illumination, and depth information is possible to prevent the attack of non-genuine operator's information such as facial image. Without memory when user uses, more will not be borrowed, usurp and lose. Therefore this utility model can promote the processing safety of ATM.
Face depth finding:
Visible ray depth finding is as it is shown on figure 3, divide two steps:
Step one: the Stereo matching of non-uncalibrated image.
By Fig. 3, visible light sensor 11 obtain the image of visible ray face. (this utility model uses relevant matches and iteration to utilize classical matching process, may be used without additive method) find initial matching point pair, then with the algorithm (this utility model is intermediate value coupling) of robust, the error hiding data dump in set is fallen by initial matching point. Recover, to polar curve set, to utilize and be newly calculated epipolar-line constraint with remaining matching double points. In special image processing unit 2, finally mate again, to obtain last matching relationship.
Step 2, according to parallax, obtains picture depth.
Infrared structure optical depth detects, as shown in Figure 4, Luo Qi grating is utilized to produce square by infrared projection unit 13, project to tested face, infrared image sensor catches square, utilizes Fourier Transform Profilomery method (Fouriertransformprofilometry, FTP), light field is carried out Fourier analysis and filtering, from the infrared stripes of deformation, extracts face three-dimensional information.
Visible ray face imaging and recognition of face performance thereof are had a great impact by the change of ambient lighting, for reducing or eliminating this impact, this patent utilizes the facial image that infrared sensor 12 shoots, depth image be with infrared sensor 12 infrared projection unit 13 to shoot infrared projection information, calculate through 3-D view processing unit 2, generate depth information figure;
Thermal infrared facial image is carried out dimensionality reduction by the dimensionality reduction technology first combined with principal component analysis (PCA) and linear judgment analysis (LDA), bulk redundancy information in image can be removed by the feature extracting thermal infrared facial image to levying extraction, only retains in image identifying significant information. then by the training of characteristic vector BP neutral net and identification. Training is to complete on other machines, and the network parameter on this device is fixing. . the minimizing of human face data amount can reduce neutral net complexity, improves training effectiveness and convergency factor. This utility model finally takes the method for Decision fusion, is weighted by summation and maximum combination, and the weight of weighting is also trained on other machines. The recognition of face information obtained by PCA and LDA recognition result and confidence level being carried out Decision fusion, provides final recognition result, Fig. 5 merges decision model.
Training set shown in Fig. 5 and feature database training can select to realize on other machines in advance. The confidence level merged also is by training acquisition. The fixing value pre-set is adopted in server in this device.
Application example:
This utility model can be applicable to desktop computer, notebook computer, ATM withdrawal machine, cell phone, access control system etc. This utility model can be used for gate inhibition, computer login, work attendance, personal identification, kindergarten ensuring pick-up safety etc. and smart city, Smart Home association area. This patent is suitable in the system that safety certification is higher, relative to ordinary two dimensional visible ray face identification system more robust, safety.
Above in conjunction with accompanying drawing, detailed description of the invention of the present utility model is described in detail, but this utility model is not restricted to above-mentioned embodiment, without departing from the spirit and scope situation of claims hereof, those skilled in the art may be made that various amendment or remodeling.
Claims (5)
1. the ATM antifraud device based on three-dimensional face identification, it is characterised in that include Face datection unit and the processing unit being connected with described Face datection unit:
Described Face datection unit arranges and is embedded in ATM, including the visible light sensor being arranged side by side successively, infrared sensor and infrared projection unit; Wherein, described visible light sensor is for gathering the visible ray facial image of user; Described infrared sensor and infrared projection unit are for coordinating the three dimensional depth facial image and infrared face image gathering user;
Described processing unit is for receiving the visible ray facial image of described user of described Face datection unit unit output, three dimensional depth facial image and infrared face image, extract characteristic of correspondence information, described characteristic information is sent to background server carry out and the characteristic information comparison of the corresponding user prestored, and receives the comparison result that background server returns.
2. ATM antifraud device according to claim 1, it is characterised in that described Face datection unit includes a substrate, and described visible light sensor, infrared sensor and infrared projection unit are installed on the substrate.
3. ATM antifraud device according to claim 2, it is characterized in that, described visible light sensor has 2, and 2 described visible light sensors are symmetrically positioned in two ends, described substrate surface left and right, and described infrared sensor and infrared projection unit are between 2 described visible light sensors.
4. the ATM antifraud device according to any one of claim 1-3, it is characterised in that also include:
Display unit, for when described processing unit receives the comparison result that described background server returns, by the display output of described comparison result.
5. the ATM antifraud system based on three-dimensional face identification, it is characterised in that include the ATM antifraud device based on three-dimensional face identification described in server and any one of claim 1-4;
The described ATM antifraud device based on three-dimensional face identification sends described characteristic information to described server;
Described server is by the characteristic information comparison of described characteristic information with the corresponding user prestored, and returns comparison result to the described ATM antifraud device based on three-dimensional face identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201521126045.XU CN205318544U (en) | 2015-12-30 | 2015-12-30 | Device and system are prevented cheaing by ATM based on three dimensional face identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201521126045.XU CN205318544U (en) | 2015-12-30 | 2015-12-30 | Device and system are prevented cheaing by ATM based on three dimensional face identification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN205318544U true CN205318544U (en) | 2016-06-15 |
Family
ID=56198980
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201521126045.XU Active CN205318544U (en) | 2015-12-30 | 2015-12-30 | Device and system are prevented cheaing by ATM based on three dimensional face identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN205318544U (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513221A (en) * | 2015-12-30 | 2016-04-20 | 四川川大智胜软件股份有限公司 | ATM (Automatic Teller Machine) cheat-proof device and system based on three-dimensional human face identification |
| CN107844744A (en) * | 2017-10-09 | 2018-03-27 | 平安科技（深圳）有限公司 | With reference to the face identification method, device and storage medium of depth information |
| CN109286757A (en) * | 2017-07-19 | 2019-01-29 | 富士施乐株式会社 | Image processing device and image processing method |
| US11170204B2 (en) | 2018-04-12 | 2021-11-09 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | Data processing method, electronic device and computer-readable storage medium |
-
2015
- 2015-12-30 CN CN201521126045.XU patent/CN205318544U/en active Active
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513221A (en) * | 2015-12-30 | 2016-04-20 | 四川川大智胜软件股份有限公司 | ATM (Automatic Teller Machine) cheat-proof device and system based on three-dimensional human face identification |
| CN105513221B (en) * | 2015-12-30 | 2018-08-14 | 四川川大智胜软件股份有限公司 | A kind of ATM machine antifraud apparatus and system based on three-dimensional face identification |
| CN109286757A (en) * | 2017-07-19 | 2019-01-29 | 富士施乐株式会社 | Image processing device and image processing method |
| CN109286757B (en) * | 2017-07-19 | 2021-12-21 | 富士胶片商业创新有限公司 | Image processing apparatus, image processing method, and program |
| US11533444B2 (en) | 2017-07-19 | 2022-12-20 | Fujifilm Business Innovation Corp. | Image processing device |
| CN107844744A (en) * | 2017-10-09 | 2018-03-27 | 平安科技（深圳）有限公司 | With reference to the face identification method, device and storage medium of depth information |
| US11170204B2 (en) | 2018-04-12 | 2021-11-09 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | Data processing method, electronic device and computer-readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105513221B (en) | A kind of ATM machine antifraud apparatus and system based on three-dimensional face identification | |
| WO2018129687A1 (en) | Fingerprint anti-counterfeiting method and device | |
| Uhl et al. | Iris-sensor authentication using camera PRNU fingerprints | |
| US8655026B2 (en) | Robust biometric feature extraction with and without reference point | |
| Mane et al. | Review of multimodal biometrics: applications, challenges and research areas | |
| CN205318544U (en) | Device and system are prevented cheaing by ATM based on three dimensional face identification | |
| JP6910236B2 (en) | Personal information registration system and personal information registration method | |
| Malathi | An integrated approach of physical biometric authentication system | |
| WO2011095041A1 (en) | Security card, system for authenticating card, financial device including the system and authentication method | |
| CN108491768A (en) | The anti-fraud attack method of corneal reflection face authentication, face characteristic Verification System | |
| Hooda | ATM security | |
| KR20180057167A (en) | An Unmanned Financial Transactions System and A Financial Transactions Method Using The Same | |
| Oruh | Three-factor authentication for automated teller machine system | |
| Malviya | Face recognition technique: Enhanced safety approach for ATM | |
| Fernandez et al. | Protection of online images against theft using robust multimodal biometric watermarking and T-norms | |
| CN112862491A (en) | Face payment security method and platform based on security unit and trusted execution environment | |
| Ashiba et al. | RETRACTED ARTICLE: Implementation face based cancelable multi-biometric system | |
| CN113033291B (en) | Face recognition method and system based on RFID | |
| Ameh et al. | Securing cardless automated teller machine transactions using bimodal authentication system | |
| Betab et al. | Fingerprints in automated teller Machine-A survey | |
| Noor | A new algorithm for minutiae extraction and matching in fingerprint | |
| Karale et al. | Online Transaction Security Using Face Recognition: A Review | |
| Patila et al. | Multi-banking ATM system services using biometrics | |
| Zhang et al. | The application of biometric identification technology in ATM system | |
| Shoukat et al. | Fraud detection system using facial recognition based on Google Teachable Machine for banking applications |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |