WO2013020577A2

WO2013020577A2 - Body part guidance control method for non contact biometric authentication device, and biometric authentication device

Info

Publication number: WO2013020577A2
Application number: PCT/EP2011/063578
Authority: WO
Inventors: Bertrand NDZANA NDZANA; Hadrien MOTTAZ; Patrick Gerber; Jean Rossier; Héctor SATIZÁBAL
Original assignee: Sensometrix Sa
Priority date: 2011-08-07
Filing date: 2011-08-07
Publication date: 2013-02-14
Also published as: WO2013020577A3

Abstract

A body part guidance control method of a non-contact biometric authentication device (2), which captures images of a body part (12) for contactless authentication, the method comprising the steps of: displaying, on a display unit, a target as a first simple geometric shape (20) at a predetermined location representative of a location for correct biometric data extraction from the body part; capturing an image of the body part within a capture plane by means of an image capture unit (4); extracting localization data of the body part with respect to the image capture unit; displaying, on the display unit (8), the body part image capture as a second simple geometric shape (22) at a variable location based on the body part localization data; providing a first guidance indication within the capture plane through the relative position between the first and second shapes and providing a second guidance indication in a perpendicular direction to the capture plane through the relative size between the first and second shapes; locking the second shape with respect to the first shape when the body part localization data of the image captured is in the vicinity of the correct location for biometric data extraction from the body part; extracting biometric data from the locked body part image capture; authenticating the extracted biometric data from the body part with previously registered body part biometric data in a storage unit (10).

Description

BODY PA RT G U I DANCE CONT ROL M ETHOD FO R NON

CONTACT B IOM ET R IC AUTH E NT ICAT ION D EVICE , AN D

B IOM ET R IC AUTH E NT ICAT ION D EV ICE

TECHNICAL FIELD

The present invention generally relates to a body part guidance control method for a non-contact biometric authentication device which performs individual authentication utilizing characteristics of a body part which is a portion of a human body, and to the device itself, and in particular relates to such a method and device in which, when detecting biometric information without contact in order to perform authentication against previously registered biometric data, the body part to be detected is guided into a detection region.

BACKGROUND OF THE INVENTION

Such body part guidance control method and device are know in the art, for instance from the document EP 1 903 509. This document describes a biometric authentication device which identifies characteristics of the body part such as a hand from captured images of the body part and performs individual authentication. The device guides a user, at the time of verification, to the image capture state at the time of registration of biometric characteristic data. At the time of registration of biometric characteristic data, hand image capture state data is extracted from an image captured by an image capture unit and is registered in a storage unit and at the time of verification the registered image capture state data is read from the storage unit and is compared with image capture state data extracted at the time of verification, and guidance of the hand is provided. Alternatively, an outline of the hand at the time of registration, taken from image capture state data at the time of registration, is displayed. It is further provided to display guidance messages relative to hand position shift, hand inclination shift, or finger's position.

Nevertheless, such body part guidance control method and device present several drawbacks. First it provides a guidance screen representing a hand silhouette corresponding to the user's hand meaning that the method and device are limited to verification systems where the user is known prior to authentication of his hand as a genuine. Further, displaying the actual hand image capture in superimposition with the registered hand silhouette raises difficulties to obtain a clean superimposed image of both displayed hands transforming the capture process into a hand contortion exercise. In addition, use of written guidance messages turned out to be not intuitive to guide the user for correct positioning of his hand and even disturbing with respect to the hand capture image and silhouette representations.

Although, more intuitive guidance is provided in the prior art, for instance in the document US 2007/0172122, it relates to image processing apparatus and method which are dedicated to objects containing a two dimensional code such as a bar code and as such are not be applicable to living body capture. Indeed, such code capture necessitates the provision of a plurality of markers on the objects in order to be able to position the captured image with respect to the capture unit which cannot be considered for biometric authentication method and device. In addition, the guidance method disclosed in US 2007/0172122 provides to display several targets one for positioning and one for sizing the code of the object which renders the code capture recognition less intuitive and more difficult. SUMMARY OF THE INVENTION

One goal of the present invention is to overcome the aforecited drawbacks by providing a simple and intuitive body part guidance control method of a non-contact biometric authentication device, which captures images of a body part for contactless authentication, and a corresponding non-contact biometric authentication device. Indeed contactless authentication is more and more recommended for health and safety reasons while not needing any mechanical support for the body part.

For that purpose, according to a first aspect, the invention concerns a body part guidance control method of a non-contact biometric authentication device, which captures images of a body part for contactless authentication, the method comprising the steps of: displaying, on a display unit, a target as a first simple geometric shape at a predetermined location representative of a location for correct biometric data extraction from the body part; capturing an image of the body part within a capture plane by means of an image capture unit; extracting localization data of the body part with respect to the image capture unit; displaying, on the display unit, the body part image capture as a second simple geometric shape similar to the first shape at a variable location based on the body part localization data; providing a first guidance indication within the capture plane through the relative position between the first and second shapes and providing a second guidance indication in a perpendicular direction to the capture plane through the relative size between the first and second shapes; locking the second shape with respect to the first shape when the body part localization data of the image captured is in the vicinity of the correct location for biometric data extraction from the body part; extracting biometric data from the locked body part image capture; authenticating the extracted biometric data from the body part with previously registered body part biometric data in a storage unit. Such guidance control method guides the user to place his hand in the correct location in a very intuitive and simple way by using only relative size and position of two simple geometric shapes such as a circle, triangle, quadrangle, etc.. Thus, there is no need to add any voice or written messages to help the user. Further, when the device is ready for acquiring, the lock position is self explaining so that the user immediately understands that he has to hold on his hand for a few moments for authentication.

According to another embodiment, it further comprises the steps of: defining a pattern of a body part; extracting state data from the body part image capture; comparing extracted state data with the defined pattern and validating the body part image capture as a body part able to be authenticated as a result of the comparison; and displaying a third guidance indication representative of the body part image capture validation. The pattern may be defined either by way of registering the body part in the storage unit or by programming characteristics representative of the body part to be checked. Such body part recognition or validation ensures that only genuine body parts lead to an authentication process. This saves processing time and gives clear indication to the user that no authentication process will be operated.

According to another embodiment, it further comprises the steps of displaying a fourth guidance indication representative of the inclination of the body part with respect to the image capture unit based on the extracted state data.

According to another embodiment, it further comprises the steps of displaying additional guidance indications is done by means of a bar light progressively lighting. Alternatively, the additional guidance indications may be displayed by any other suitable displaying means such as an LCD, another guidance screen or color LEDs, etc.

According to another embodiment, it further comprises the steps of displaying the fourth guidance indication is done by way of deformation of the second shape relatively to the first shape. According to another embodiment, the extracted state data comprises data relative to a shape of the body part.

According to another embodiment, the locking step may consist in comparing a center of mass location of the body part image capture with the correct location for biometric data extraction from the body part defined in the capture plane center. Alternatively or additionally, the locking step may be depending on other criteria such as checking the size, conformity or recognition of the detected body part.

According to another embodiment, the extracted biometric data is a vein pattern of a hand.

According to a second aspect, the invention concerns a non-contact biometric authentication device, which verifies registered biometric data for a user against biometric data detected from a body part to perform individual authentication, comprising: a storage unit, which stores previously registered biometrics data of body parts; an image capture unit, which captures, without contact, an image of the body part within a capture plane; a processing unit, which extracts localization data of the body part with respect to the image capture unit; a display unit, which displays a target as a first simple geometric shape at a predetermined location representative of a location for correct biometrics data extraction from the body part and the body part image capture as a second simple geometric shape similar to the first shape at a variable location based on the body part localization data; wherein the display unit further provides a first guidance indication within the capture plane through the relative position between the first and second shapes and a second guidance indication in a perpendicular direction to the capture plane through the relative size between the first and second shapes; and wherein the processing unit locks the second shape with respect to the first shape when the body part localization data of the image captured is in the vicinity of the correct location for biometrics data extraction from the body part; extracts biometrics data from the locked body part image capture; and authenticates the extracted biometrics data from the body part with the previously registered body part biometrics data stored in storage unit.

According to a third aspect, the invention concerns a computer readable medium storing a program causing a computer to execute image processing, the image processing comprising the steps of the body part guidance control method of a non-contact biometric authentication device according to the first aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

Other features and advantages of the invention will appear upon reading the following description, given by way of example only, which refers to the annexed drawings in which:

- Figure 1 schematically represents a biometric acquisition platform according to an embodiment of the present invention,

- Figure 2 schematically represents the main steps of the guidance method,

- Figures 3a to 3d represents the guidance screen according to an embodiment of the invention,

- Figure 4 represents the steps of the guidance control method according to another embodiment,

- Figure 5 represents the steps of the guidance control method according to another embodiment. DETAILED DESCRIPTION OF THE INVENTION

Figure 1 schematically represents a non contact biometric authenticating device 2 or acquisition platform comprising an image capture unit 4, a processing unit 6, a display unit 8 and a storage unit 10.

The image capture unit 4 may comprise a camera, several embedded infrared LEDs and a proximity sensor that detects when something is in front of the camera at a distance smaller than a certain threshold. This presence triggers the acquisition process that is executed up to the point when a body part such as a hand 12 as been successfully recognized in front of the camera at a distance range compatible with the subsequent identification or authentication process. An image is captured from the camera sensor within an integrated circuit by flashing the hand 12 with the embedded infrared lights.

The processing unit 6 is arranged for extracting several kinds of data from the captured image. The processing unit extracts state, localization and biometric data from the captured image. State data may be of any kind that allows when compared to a predefined pattern to acknowledge the presence of a hand 12 in front of the camera. For instance, the state data may comprise different parameters such as the shape of the hand or the fingers. The processing unit further extracts localization data to determine whether the hand is in a right location with respect to the image capture unit 4. Finally, when the hand 12 is considered to be in a right location for image capture, the processing unit extracts biometric data such as palm vein pattern revealed from the captured image of the hand taken with infrared lights flashing at a power tailored to a correct exposure.

The display unit 8, controlled by the processing unit 6, displays several guidance indications to the user in order to let him position correctly his hand 12 with respect to the image capture unit 4. The guidance indications will be detailed later in relation with Figures 3a-3d. The storage unit 10 is used to store previously registered biometric data corresponding to the different users already registered in the system. These previously registered biometric data will be used to determine if the palm vein pattern extracted from the current captured image corresponds to one of the registered users so as to authenticate him. It is to note that the storage unit have not to be integrated within the authentication device but could well be a remote storage unit.

Figure 2 schematically represents the main steps of the guidance method. As already mentioned above, the authentication device will first process with the image capture of the hand within a capture plan 18 and then guide the user in order to optimize the position of his hand 12 to be captured by the image capture unit 4. The display unit 8 preferably comprises a guidance screen 14 and a bar light 16 for example made of LEDs.

Figures 3a to 3d represents the display unit in more details. Indeed, in order to guide the user and facilitate a good placement of his hand in front of the camera at the desired distance, we propose to use a human machine interface made of a guidance screen 14 along with a progress bar or bar light 16.

The guidance screen 14 mainly displays a first simple geometric shape such as a first circle 20 which represents the target representative of a location for correct biometric data extraction from the hand, said target being preferably stationary on the screen and a second simple geometric shape 22 similar to the first one, even though a different shape may be considered, which position and size on the screen depend on the hand localization data, i.e. the actual symbolic position of the user hand. Its center position may be given the position of the center of mass of all the binarized pixels of the captured image of the hand. Thus, the position of the second circle 22 relatively to the target circle 20 indicates to the user in which direction he should move his hand within the capture plane 18. In figure 3a, the user should move his hand forward and left, while in figure 3b, he should move a bit backward and left and in figure 3, a bit forward and right. The size of the second circle 22, i.e. its radius, is dependent on the distance of the hand to the camera. In other words, when the hand is near the camera, the amount of binarized pixels of the captured image is quite great and so is the radius of the second circle 22 (figure 3c) indicating to the user that he should move his hand away from the camera along a perpendicular axis 24 to the capture plane 18. On the other hand, when the hand is too far away from the camera, the amount of binarized pixels is small, and so does the radius of the second circle (figures 3a and 3b) indicating to the user that he should move closer to the camera along the axis 24. Finally, when the number of binarized pixels lies in a pre-defined range indicating that the hand in front of the camera is at a good distance, the radius of the second circle 22 is preferably just below the radius of the target circle 20. If in addition, the position of the center of mass of the second circle is not too far away from the center of the capture window, then the hand vein image is considered as valid and is acquired. Acquisition of the hand may be linked to additional criteria such as the size, conformity or recognition of the hand. When all criteria are validated, the second circle 22 then jumps in the middle of the first circle 20, as represented in Figure 3d, and locks that position while the authentication is done by the system. This behaviour (Figure 3d) informs the user that his hand has been correctly captured and that he has to wait a bit to be recognized.

The progress bar 16 indicates the progression through hand recognition tests as well as other potential tests such as the inclination of the hand with respect to the image capture unit. The progress bar usually increases light by illuminating the corresponding LEDs, as shown in Figure 3a-3d, as the user succeeds the hand positioning tests. Of course alternatively, any other suitable displaying means may be used instead of a LED bar. Figure 4 represents the steps of the guidance control method according to a first embodiment. As a preliminary step, biometric data of different users are registered in the storage unit. This preliminary step may be done anytime prior to the authentication process that will be described hereafter. The acquisition of biometric data during registration is done in the same way as for the authentication process as explained below. During step 1 , the guidance screen displays the target circle at a predetermined location representative of a location for correct biometric data extraction from the hand. During step 2, the image capturing unit captures the hand image. During step 3, the processing unit extracts localization data of the hand with respect to the image capture unit. At step 4, the guidance screen displays the second circle representative of the actual position of the hand with respect to the image capturing unit. If the localization data is neither correct nor enough reliable, then the process goes back to step 2. During step 5, the processing unit locks the second shape with respect to the first shape when the localization data of the image captured is in the vicinity of the correct location for biometric data extraction. Then at step 6, the processing unit extracts biometric data from the hand and at step 7 authenticates the extracted biometric data with the previously registered biometric data stored in the storage unit.

Figure 5 represents the steps of the guidance control method according to a second embodiment. This second embodiment differs from the first as follows. Preliminary step further comprises registration of a hand pattern in storage unit. Step 3 further comprises extraction of hand state data. Further to step 3, at step 31 , the processing unit compares the extracted state data with the registered pattern and at step 32 validates the hand image capture as a hand able to be authenticated and at step 33 the progress bar indicates the recognition of a hand to be authenticated. At step 4, if any of the validation test or localization data is neither correct nor enough reliable, then the process goes back to step 2.

Having described the invention with regard to certain specific embodiments, it is to be understood that these embodiments are not meant as limitations of the invention. Indeed, various modifications, adaptations and/or combination between embodiments may become apparent to those skilled in the art without departing from the scope of the annexed claims. In particular, the simple geometric shapes in Figures 3a-3d have been represented in the shape of a circle, but could also be any other simple and intuitive representation such a triangle, rectangle, quadrangle, ring, disk, etc.

Claims

CLA I MS

1 . A body part guidance control method of a non-contact biometric authentication device (2), which captures images of a body part (12) for contactless authentication, the method comprising the steps of:

displaying, on a display unit, a target as a first simple geometric shape (20) at a predetermined location representative of a location for correct biometric data extraction from the body part;

capturing an image of the body part within a capture plane by means of an image capture unit (4);

extracting localization data of the body part with respect to the image capture unit;

displaying, on the display unit (8), the body part image capture as a second simple geometric shape (22) at a variable location based on the body part localization data; providing a first guidance indication within the capture plane through the relative position between the first and second shapes and providing a second guidance indication in a perpendicular direction to the capture plane through the relative size between the first and second shapes;

locking the second shape with respect to the first shape when the body part localization data of the image captured is in the vicinity of the correct location for biometric data extraction from the body part;

extracting biometric data from the locked body part image capture; authenticating the extracted biometric data from the body part with previously registered body part biometric data in a storage unit (10).

2. The body part guidance control method of a non-contact biometric authentication device (2) according to claim 1 , further comprising the steps of:

defining a pattern of a body part (12); extracting state data from the body part image capture;

comparing extracted state data with the defined pattern and validating the body part image capture as a body part able to be authenticated as a result of the comparison; and

displaying a third guidance indication representative of the body part image capture validation.

3. The body part guidance control method of a non-contact authentication device (2) according to claim 2, further comprising the step of displaying a fourth guidance indication representative of the inclination of the body part with respect to the image capture unit based on the extracted state data.

4. The body part guidance control method of a non-contact authentication device (2) according to claim 2 or 3, wherein the step of displaying additional guidance indications is done by means of a bar light (16) progressively lighting.

5. The body part guidance control method of a non-contact authentication device (2) according to 3, wherein the step of displaying the fourth guidance indication is done by way of deformation of the second shape relatively to the first shape.

6. The body part guidance control method of a non-contact authentication device (2) according to any of claims 2 to 5, wherein the extracted state data comprises data relative to a shape of the body part.

7. The body part guidance control method of a non-contact authentication device (2) according to any of claims 1 to 6, wherein the locking step consists in comparing a center of mass location of the body part image capture with the correct location for biometric data extraction from the body part defined in the capture plane center.

8. The body part guidance control method of a non-contact authentication device (2) according to any of claims 1 to 7, wherein the extracted biometric data is a vein pattern of a hand.

9. A non-contact biometric authentication device (2), which verifies registered biometric data for a user against biometric data detected from a body part (12) to perform individual authentication, comprising:

a storage unit (10), which stores previously registered biometrics data of body parts;

an image capture unit (4), which captures, without contact, an image of the body part within a capture plane (18);

a processing unit (6), which extracts localization data of the body part with respect to the image capture unit;

a display unit (8), which displays a target as a first simple geometric shape (20) at a predetermined location representative of a location for correct biometrics data extraction from the body part and the body part image capture as a second simple geometric shape (22) at a variable location based on the body part localization data;

wherein the display unit further provides a first guidance indication within the capture plane through the relative position between the first and second shapes and a second guidance indication in a perpendicular direction to the capture plane through the relative size between the first and second shapes;

and wherein the processing unit locks the second shape with respect to the first shape when the body part localization data of the image captured is in the vicinity of the correct location for biometrics data extraction from the body part; extracts biometrics data from the locked body part image capture; and authenticates the extracted biometrics data from the body part with the previously registered body part biometrics data stored in storage unit.

10. The non-contact biometric authentication device (2) according to claim 9, further comprises a defined pattern of a body part; wherein the processing unit (6) extracts state data from the body part image capture and compares extracted state data with the defined pattern, validates the body part image capture as a body part able to be authenticated as a result of the comparison; and displays a third guidance indication representative of the body part image capture validation.

1 1 . The non-contact biometric authentication device (2) according to claim 10, wherein the processing unit (6) further displays a fourth guidance indication representative of the inclination of the body part with respect to the image capture unit based on the extracted state data.

12. The non-contact biometric authentication device (2) according to claim 10 or 1 1 , further comprises a bar light (16) which displays additional guidance indications by progressive lighting.

13. The non-contact biometric authentication device (2) according to any of claims 9 to 12, wherein the extracted biometric data is a vein pattern of a hand.

14. A computer readable medium storing a program causing a computer to execute image processing, the image processing comprising the steps of the body part guidance control method of a non-contact biometric authentication device according to any of claims 1 to 8.