The present invention contains subject matter related to Japanese Patent Application JP2007-046123 filed in the Japanese Patent Office on Feb. 26, 2007, the entire contents of which being incorporated herein by reference.
1. Field of the Invention This invention relates to an image determination device, an image determination method, and a program, which are desirably applied to the biometric authentication.
2. Description of the Related Art
As one of biometric authentication subjects, a blood vessel has been employed. In the blood vessel, the deoxygenated hemoglobin (venous blood) and oxygenated hemoglobin (arterial blood) are provided with properties of specifically absorbing light in the near-infrared light band (near infrared ray), and, by utilizing the properties, an image of a blood vessel of a finger is picked up.
As a method to direct a finger toward the image pickup range, there has been suggested a method in which an image of a finger placed on a finger placement table is picked up, and, with the first joint and second joint is from the fingertip in the picked up image set as the criteria, it is determined whether or not the finger is set within the image pickup range, and a guidance to pick up an image of the finger again is displayed according to the determination result (for example, refer to Patent Document 1; Registration of Utility Model Mo. 3100993).
Meanwhile, in case of obtaining a constant resolution, setup of the image pickup range becomes an important factor in terms of downsizing. In this regard, as for the above-described direction method, in determining whether or not the finger is set on the suitable position within the image pickup range by using an image, since the existence of the first joint and second joint is set as the criteria, there is a problem that an image pickup range which is equal to or more than a range which can set the finger from the first joint to the second joint thereof needs to be arranged.
In view of the above-identified circumstances, it is therefore desirable to provide an image determination device, an image determination method, and a program, which can improve the degree of freedom in setting the image pickup range.
According to an embodiment of the present invention, there is provided an image determination device including: an extraction means for extracting a figuration pattern of an identification subject in a finger reflected on an image; a detection means for detecting a joint line in the finger; a determination means for determining the image as an image to be registered or an image to be collated with a registration subject when a joint line exists in a first range which is set in one of respective regions obtained when separating the image with the center line that corresponds to a direction perpendicular to the longitudinal direction of the finger and that is set as the border, and a blood vessel amount which is equal to or more than a predetermined threshold value exists in a second range which is set in the other region different from the region in which the first range is set.
According to an embodiment of the present invention, there is also provided an image determination method including: a first step of setting a first range in one of respective regions obtained when separating an image with the center line that corresponds to a direction perpendicular to the longitudinal direction of a finger reflected on the image and that is set as the border, and, in case a joint line of the finger does not exist in the set first range, setting the first range in the other region of the respective regions; a second step of, when the joint line exists in the first range, setting a second range in the other region different from the region in which the first range is set; and a third step of, when the joint line exists in the first range, and a blood vessel amount which is equal to or more than a predetermined threshold value exists in the second range, determining the image as an image to be registered or an image to be collated with a registration subject.
According to an embodiment of the present invention, there is also provided a program that makes a control unit controlling a work memory execute: setting a first range in one of respective regions obtained when separating an image input to the control unit with the center line that corresponds to a direction perpendicular to the longitudinal direction of a finger reflected on the image and that is set as the border, and, in case a joint line of the finger does not exist in the set first range, setting the first range in the other region of the respective regions; when the joint line exists in the first range, setting a second range in the other region different from the region in which the first range is set; and when the joint line exists in the first range, and a blood vessel amount which is equal to or more than a predetermined threshold value exists in the second range, determining the image as an image to be registered or an image to be collated with a registration subject.
As described above, according to the present invention, when a joint line exists in a predetermined position (first range) set in one region side of an image, it can be determined whether or not the joint line is a line corresponding to the second joint of a finger according to the degree of a blood vessel amount in a predetermined position (second range) set in the other region side.
Accordingly, even if the image pickup range does not have a range in which part of a finger from the first joint to the second joint is set in, the authentication device 1 can determine whether or not a region near the second joint between the first joint and the second joint is located at the center of an image by the existence of a joint line, and thus, an image determination device, an image determination method, and a program capable of improving the degree of freedom in setting the image pickup range can be realized.
The nature, principle and utility of the present invention will become more apparent from the following detailed description when read in conjunction with the accompanying drawings in which like parts are designated by like reference numerals or characters.
In the accompanying drawings:
Now, embodiments of the present invention will be described in greater detail by referring to the accompanying drawings.
The control unit 10 is configured as a computer that includes a central processing unit (CPU) which controls the entire authentication device 1, a read only memory (ROM) which stores various programs and setup information, and a random access memory (RAM) which works as a work memory for the CPU.
To the control unit 10, an execution command COM1 for a mode (referred to as blood vessel registration mode, hereinafter) to register a blood vessel of a user to be registered (referred to as registrant, hereinafter), or an execution command COM2 for a mode (referred to as authentication mode, hereinafter) to determine the existence or absence of the registrant is input from the operation unit 11 according to the user operation.
The control unit 10 determines the mode to he executed based on the execution commands COM1, COM2, and, based on a program corresponding to the determination result, arbitrarily controls the image pickup unit 12, memory 13, interface 14, and notification unit 15, and execute the blood vessel registration mode or authentication mode.
The image pickup unit 12 has a camera which sets a space over an area of the housing of the authentication device 1, on which a finger is placed, as the image pickup space, and adjusts the lens position of the optical system or the camera, the diaphragm value of the diaphragm, and the shutter speed (exposure time) of the image pickup element using an exposure value (EV) set up by the control unit 10 as a criterion.
The image pickup unit 12 has a near infrared light source that irradiates a near infrared ray to the image pickup space, and turns on the near infrared light source for a time period specified by the control unit 10, and picks up images of an image pickup subject which are reflected on the image pickup surface of the image pickup element every predetermined cycle, and outputs image data related to images generated as the image pickup result to the control unit 10 in series.
The memory 13 is, for example, a flash memory, and stores or reads out data specified by the control unit 10.
The interlace 14 sends and receives various data to and from an external device connected to the authentication device 1 through a predetermined transmission path.
The notification unit 15 includes a display unit 15a and an audio output unit 15b, and the display unit 15a displays characters and figures based on display data sent from the control unit 10 on a display screen, while the audio output unit 15b outputs audio based on audio data sent from the control unit 10 from a speaker.
Next, the blood vessel registration mode will be described. When determining the blood vessel registration mode as a mode to be executed, the control unit 10 sets the operation mode to the blood vessel registration mode, and makes the notification unit 15 notify that a finger has to be arranged in the image pickup space.
At this time, the control unit 10 makes the camera arranged in the image pickup unit 12 pick up images, and turns on the near infrared light source arranged in the image pickup unit 12.
In this state, when a finger is arranged in the image pickup space, a near infrared ray which is irradiated from the near infrared light source and passes through the inside of the finger goes to the image pickup element through the optical system and diaphragm of the camera as light which projects a blood vessel, and an image of the blood vessel arranged in the inside of the finger is reflected on the image pickup surface of the image pickup element. Accordingly, in the image based on image data which is generated as the image pickup result by the image pickup unit 12, the blood vessel is reflected.
The control unit 10 performs predetermined image processing for the image data input from the image pickup unit 12 to generate data to be identified (referred to as identification data, hereinafter), and makes the memory 13 store the identification data for registration.
In this way, the control unit 10 can execute the blood vessel registration mode.
Next, the authentication mode will be explained. When determining the authentication mode as a mode to be executed, the control unit 10 sets the operation mode to the authentication mode, and makes the notification unit 15 notify that a finger has to be arranged in the image pickup space, and makes the camera arranged in the image pickup unit 12 pick up images, and turns on the near infrared light source.
The control unit 10 performs the same image processing as that performed in the blood vessel registration mode for the image data input from the image pickup unit 12 to generate identification data. Then, the control unit 10 collates thus generated identification data and the identification data stored in the memory 13, and, according to the degree of data correlation which is obtained as the collation result, determines whether or not a person can be approved as the registrant.
In this case, when it is determined that the person is unable to be approved as the registrant, the control unit 10 notifies the person of the disapproval visually as well as aurally through the display unit 15a and the audio output unit 15b. On the other hand, in case it is determined that the person can be approved as the registrant, the control unit 10 sends data indicative of the approval as the registrant to a device connected to the interface 14. With the data indicative of the approval as the registrant set as a trigger, the device connected to the interface 14 carries out predetermined processing, such as locking a door for a predetermined time period, or releasing a restricted operation mode. The processing should be executed when the authentication is successfully performed.
In this way, the control unit 10 can execute the authentication mode.
Next, image processing in the control unit 10 will be described. As shown in
The pattern extraction unit 21 extracts a figuration pattern of a blood vessel reflected on an image represented by image data D1 input from the image pickup unit 12, and sends image data D2 related to an image in which the figuration pattern of the blood vessel is extracted to the noise elimination unit 22 and the joint line detection unit 23.
One example of the extraction method in the pattern extraction unit 21 will be explained. As preprocessing, the pattern extraction unit 21 highlights the contour reflected on the image, using a differentiation filter such as a Gaussian filter, a Log filter, etc. Furthermore, as preprocessing, the pattern extraction unit 21 corrects and rotates the image which has its contour highlighted such that the contour along the longitudinal direction of the finger comes to be parallel with the vertical direction (up and down direction) of the image, and cuts out an image of a region of a predetermined size with the center position set to the criterion from thus corrected and rotated image.
In this state, the pattern extraction unit 21 converts thus cut out image to a binary image with a set luminance value used as the criterion, and extracts the figuration pattern of the blood vessel as a line (referred to as blood vessel line, hereinafter) by detecting the center of width or the luminance peak of the width of a part (object) corresponding to the blood vessel reflected on the binary image.
Of image (binary image) represented by the image data D2 sent from the pattern extraction unit 21, the noise elimination unit 22 eliminates components (referred to as transverse wrinkle components, hereinafter) corresponding to wrinkle (referred to as transverse wrinkles, hereinafter) along a direction perpendicular to the longitudinal direction of the finger as noise, and sends image data D3 related to the image which has its transverse wrinkle components eliminated to the joint line defection unit 23 and the position determination unit 24.
Longitudinal wrinkle components on the finger surface are not eliminated since, in general, there is a tendency that transverse wrinkles are more noticeable than wrinkles (longitudinal wrinkles) along the longitudinal direction of a finger.
One example of the elimination method in the noise elimination unit 22 will be explained. For example, as shown in
Specifically, for example, as shown in
Furthermore, for example, as shown in
Moreover, for example, as shown in
In this way, the noise elimination unit 22 specifies the respective pixels as a focused pixel, and disperses the transverse wrinkle components by replacing the luminance value of the focused pixel with a luminance average value of pixels of a predetermined number continuing in the vertical direction (up and down direction) with the focused pixel set to the center, thereby eliminating the transverse wrinkle components.
Of transverse wrinkle components obtained from the difference between the image (
One example of the detection method in the joint line detection unit 23 will be explained. As shown in
Then, as shown in
Of the image represented by the image data D3 sent from the noise elimination unit 22, when a joint line of the finger exists in a first range which is set in one of respective regions obtained when separating the image with the center line corresponding to a direction perpendicular to the longitudinal direction of the finger set to the border, and a blood vessel amount which is equal to or more than a predetermined threshold value exists in a second range which is set in the other region different from the region in which the first range is set, the position determination unit 24 determines the image as an image to be registered or an image to be collated with a registration subject, and generates the image as identification data D5. The identification data D5 is registered in the memory 13 in case of the blood vessel registration mode, and is collated with identification data registered in the memory 13 in case of the authentication mode.
One example of the determination method in the position determination unit 24 will be explained. As shown in
In
Then, the position determination unit 24 recognizes a joint line from the position data D4 sent from the joint line detection unit 23, and, in case a joint line does not exist in the joint line detection range S1 (
In case a joint line does not exist in the joint line detection range S1 set in the upper region TR1 (
In this case, the notification unit 15 notifies contents that the arrangement position of the finger is largely distant from the suitable position in the image pickup range, or that the image pickup environment is inferior (
On the other hand, in case a joint line is detected in the joint line detection range S1 set in the lower region TR2 or in the upper region TR1 (
In
In case a blood vessel amount equal to or more than a predetermined threshold value does not exist in the blood vessel amount detection range S2 (
In this case, the notification unit 15 notifies contents that the arrangement position of the finger is slightly distant from the suitable position in the image pickup range (
On the other hand, in case a blood vessel amount equal to or more than a predetermined threshold value exists in the blood vessel amount detection range S2 (
In the above-described configuration, the authentication device 1 extracts a figuration pattern of a blood vessel of a finger reflected on an image (
Then, when a joint line exists in the joint line detection range S1 set up in the lower region TR2 (or upper region TR1) obtained when separated with the center line LN corresponding to a direction perpendicular to the longitudinal direction of the finger set to the border, and a blood vessel amount which is equal to or more than a predetermined threshold value exists in the blood vessel amount detection range S2 set in the upper region TR1 (or lower region TR2) (
When a joint line exists in the joint line detection range S1 located at the lower side of the image, the authentication device 1 can determine whether or not the joint line is the second joint line according to the degree of the blood vessel amount in the blood vessel amount detection range S2 located at the upper side of the image (
Accordingly, even if the image pickup range does not have a range in which part of a finger from the first joint to the second joint is set in, the authentication device 1 can determine whether or not a region near the second joint between the first joint and the second joint is located at the center of an image by the existence of a joint line, which can improve the degree of freedom in setting the image pickup range.
When the degree of freedom in setting the image pickup range is improved, since it becomes possible to flexibly correspond to a request of not forcing a person to fix a finger, a request on design, a request of reducing size, etc., the method which can determine whether or not a region near the second joint between the first joint and the second joint is located at the center of an image by the existence of a joint line becomes useful specifically.
Furthermore, the authentication device 1 in this embodiment detects a joint line from an image from which a figuration pattern of a blood vessel is extracted (
According to the above-described configuration, when a joint line exists in the joint line detection range S1 located at the lower side of the image, it is determined whether or not the joint line is the second joint line according to the degree of the blood vessel amount in the blood vessel amount detection range S2 located at the upper side of the image. Thus, even if the image pickup range does not have a range in which part of the finger from the first joint to the second joint is set in, it can be determined whether or not a region near the second joint between the first joint and the second joint is located at the center of an image by the existence of a joint line. As a result, the authentication device 1 capable of improving the degree of freedom in setting the image pickup range can be realized.
In the above-described embodiments, an extraction unit (pattern extraction unit 21) extracts a figuration pattern of a blood vessel in the inside of the finger as a line, to which the present invention is not restricted, and a figuration pattern of a blood vessel may be extracted as a point.
Specifically, after extracting a figuration pattern of a blood vessel as a line (blood vessel line), the end point, branching point, and bending point of the blood vessel line are extracted by employing an extraction method referred to as the Harris corner, or an extraction method disclosed in Japanese Patent Application No. 2007-46089.
In case of extracting a point, the noise elimination unit 22 can be omitted, and, as shown in
Furthermore, as an identification subject, in the above-described embodiment, a blood vessel in the inside of the finger is employed, to which the present invention is not restricted, and a nerve in the inside of the finger or a fingerprint on the surface of the finger may be employed. In case of employing a fingerprint, by executing the above-described image processing with respect to image data which is obtained by irradiating a near infrared ray to the finger to pick up an image thereof, effects similar to those in the above-described embodiments can be obtained.
Moreover, in the above-described embodiment, from an image which is obtained as the image pickup result by the image pickup unit 12, a figuration pattern of a blood vessel of the finger reflected on the image is extracted (
In the control method, for example, the control unit 10 estimates the distance to a blood vessel or to the surface of the finger based on the contrast or phase of an image obtained as the image pickup result by the image pickup unit 12, or sets the distance to a blood vessel or to the surface of the finger in a ROM, and shifts an optical lens in the image pickup unit 12 to a position corresponding to the distance.
In this way, when extracting a figuration pattern of a blood vessel, the figuration pattern can be extracted based on an image which has its blood vessel components highlighted in proportion to transverse wrinkle components. On the other hand, when detecting a joint line, the joint line can be detected based on an image which has its transverse wrinkle components highlighted in proportion to blood vessel components. Accordingly, both the accuracy in extracting the figuration pattern and the accuracy in detecting the joint line can be improved.
Furthermore, in the above-described embodiment, the threshold value of a blood vessel amount set with respect to the blood vessel amount detection range S2 is fixed, to which the present invention is not restricted, and the threshold value of a blood vessel amount may be variable according to the waveform state of the luminance histogram.
The relationship between the waveform state of the luminance histogram and the difficulty in reflecting a blood vessel will foe described. In general, it is known that the degree of difficulty in reflecting a blood vessel becomes different according to biological body elements such as sex, race, age, constitution of a biological body. On the other hand, the waveform state of the luminance histogram becomes different when the biological body elements are different such as a biological body in which the bone is thin and the amount of body fat is large (
Accordingly, the degree of difficulty in reflecting a blood vessel can be specified to some extent according to the pattern of the waveform state of the luminance histogram.
In case of a waveform pattern in which the degree of the difficulty in reflecting a blood vessel is large, when the threshold value of a blood vessel amount with respect to the blood vessel amount detection range S2 is small, it is wrongly determined that the blood vessel line is not the second joint due to the difficulty in reflecting a blood vessel can be reduced as compared with a case in which the threshold value is fixed, even if a blood vessel line existing in the joint line detection range S1 is the second joint.
Specifically, the position determination unit 24 obtains a luminance histogram of an image represented by the image data D1 sent from the image pickup unit 12, and, when the waveform pattern of the luminance histogram is a waveform pattern in which the difficulty in reflecting a blood vessel is large, switches the threshold value of a blood vessel amount with respect to the blood vessel amount detection range S2 to a value which is reduced by a predetermined ratio with respect to the criterion value. In this way, when a joint line exists in the joint line detection range S1, it can further be correctly determined whether or not the joint line is the second joint line.
Furthermore, in the above-described embodiment, a case in which the size of the blood vessel amount detection range S2 is fixed is described, to which the present invention is not restricted, and the size of the blood vessel amount detection range S2 may be variable according to the waveform state of the luminance histogram.
Specifically, the position determination unit 24 obtains a luminance histogram of an image represented by the image data D1 sent from the image pickup unit 12, and, when the waveform pattern of the luminance histogram is a waveform pattern in which the difficulty in reflecting a blood vessel is large, switches the size of the blood vessel amount detection range S2 so that the size becomes large with respect to the criterion value. In this way, similar to the case in which the threshold value with respect to the blood vessel amount detection range S2 is variable, when a joint line exists in the joint line detection range S1, it can further be correctly determined whether or not the joint line is the second joint line.
Furthermore, in the above-described embodiment, pixels continuing in the vertical direction (up and down direction) with the focused pixel set to the center are fixed to five pixels (
Specifically, the position determination unit 24 obtains a luminance histogram of an image represented by the imago data D1 sent from the image pickup unit 12, and, when the waveform pattern of the luminance histogram is a waveform pattern in which the difficulty in reflecting a blood vessel is large, switches the number of pixels continuing in the vertical direction (up and down direction) with the focused pixel set to the center so that the number becomes small with respect to the criterion value. In this way, it becomes possible to generate an image which has its transverse wrinkle components further smoothed.
Furthermore, in the above-described embodiment, the above-described image processing is performed in accordance with a program stored in a ROM, to which the present invention is not restricted, and the above-described image processing may be performed in accordance with a program which is installed from a program storage medium such as a Compact Disc (CD), Digital Versatile Disc (DVD), a semiconductor memory, or a program which is downloaded from a program-providing server on the Internet.
Moreover, in the above-described embodiment, the above-described image processing is executed by the control unit 10, to which the present invention is not restricted, and part of the processing may be executed by a graphics workstation.
Moreover, in the above-described embodiment, the authentication device 1 provided with the image pickup function, collation function, and registration function is employed, to which the present invention is not restricted, and there may be employed a configuration in which, according to the use application, the respective functions or part of the functions are separated to single devices.
The present invention is applicable to the field of the biometric authentication.
It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and alterations may occur depending on design requirements and other factors insofar as they are within the scope of the appended claims or the equivalents thereof.
Number | Date | Country | Kind |
---|---|---|---|
P2007-046123 | Feb 2007 | JP | national |