Blood vessel image photographing apparatus and personal identification system
The blood vessel imaging device corrects distortions in finger blood vessel patterns caused by bending by detecting brightness saturation areas, ensuring accurate authentication without additional cameras, addressing the challenge of distorted patterns and high costs in existing systems.
Patent Information
- Application Number
- JP2023087257
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-26
- Publication Date
- 2024-12-06
AI Technical Summary
Biometric authentication devices using finger blood vessel patterns face challenges in accurately authenticating users when fingers are bent, as this distorts the patterns, and existing solutions requiring multiple cameras increase manufacturing costs.
A blood vessel imaging device that captures images of a user's finger using near-infrared light, detects finger bending by identifying belt-shaped brightness saturation areas, and corrects distortions in the images to ensure accurate authentication.
Enables accurate user authentication even when fingers are bent by detecting and correcting distortions in the blood vessel patterns, without the need for additional cameras, thus reducing manufacturing costs.
Smart Images

Figure 2024170218000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a blood vessel image capturing device and an individual authentication system. [Background technology]
[0002] One type of biometric authentication device that can perform highly accurate personal authentication is one that uses blood vessel patterns. This type of biometric authentication device uses the fact that blood vessel patterns differ from person to person, and performs personal authentication by, for example, shining near-infrared light on a finger to reveal the finger vessel pattern, which is then compared with pre-registered finger vessel pattern information.
[0003] In a biometric authentication device that uses finger vein patterns, if a user attempts to authenticate a person by bending his or her finger even though the finger's vein pattern has been registered with the user's finger extended, bending the finger can distort the vein pattern, making it impossible to authenticate the person correctly.
[0004] Patent Document 1 describes that by installing two cameras in the authentication device, a distance measurement based on a general stereo vision technology is used to measure the distance between the finger surface and the device, thereby determining whether the finger is floating or bent. [Prior art documents] [Patent documents]
[0005] [Patent Document 1] Patent No. 6005750 Summary of the Invention [Problem to be solved by the invention]
[0006] The authentication device of Patent Document 1 can correctly authenticate a person by detecting and correcting finger bending, but it requires the installation of two cameras, which increases manufacturing costs.
[0007] An object of the present invention is to provide a technique that can correctly authenticate a user even when the user holds a bent finger over a blood vessel image capturing device. [Means for solving the problem]
[0008] In order to solve the above problems, one representative vascular image capturing device of the present invention is a vascular image capturing device that captures an image of the blood vessels of a user's finger, and is equipped with a light source that irradiates light onto the dorsal side of the finger, an imaging means that captures an image of the light-irradiated finger from the pad side of the finger, and a control unit that detects whether the finger is bent by detecting a band-shaped brightness saturation area that has passed through the joint of the finger from the captured vascular image. Effect of the Invention
[0009] According to the present invention, even when a user holds a bent finger over a blood vessel image capturing device, the present invention can correctly authenticate the user.
[0010] Problems, configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]
[0011] [Figure 1] 1 is a side view showing an example of the configuration of a blood vessel image capturing device according to a first embodiment. [Figure 2A] 1 is a side view showing an example of how a bent finger is imaged by the blood vessel image capturing device of the first embodiment. FIG. [Figure 2B] FIG. 13 is a diagram showing an example of a captured image when a light amount 20% weaker than the light amount at which the blood vessels in a finger are clearly visible is irradiated. [Figure 2C] FIG. 13 is a diagram showing an example of a captured image when a light amount sufficient to clearly show blood vessels in a finger is irradiated. [Figure 2D] FIG. 13 is a diagram showing an example of an image captured when a light amount 20% stronger than the amount of light that clearly shows the blood vessels in a finger is irradiated. [Figure 3A] 11 is a side view showing another example of how a bent finger is imaged by the blood vessel image capturing device of the first embodiment. FIG. [Figure 3B]11A and 11B are diagrams illustrating an example of a captured image when a light amount that does not cause a brightness saturation region other than the reflection of a reflected light source is irradiated. [Figure 3C] 11 is a diagram showing an example of a captured image when a light amount 1.5 times the amount of light at which no brightness saturated region other than the reflection of a reflected light source occurs is irradiated. FIG. [Figure 3D] 11 is a diagram showing an example of a captured image when a light amount twice the amount of light that does not cause a brightness saturation region other than the reflection of a reflected light source is irradiated. FIG. [Figure 4A] 4 is a diagram showing an example of a finger blood vessel image captured by the blood vessel image capturing device of the first embodiment when the finger is bent. FIG. [Figure 4B] 10 is a diagram showing another example of a finger blood vessel image captured by the blood vessel image capturing device of the first embodiment with a bent finger. FIG. [Figure 4C] 3 is a diagram showing an example of a finger blood vessel image corrected by the blood vessel image capturing device 100 of the first embodiment. FIG. [Diagram 5] FIG. 11 is a diagram illustrating an example of the configuration of an individual authentication system according to a second embodiment. [Figure 6] 11 is a flowchart showing an example of an identity authentication process executed by the identity authentication device of the second embodiment. [Figure 7] FIG. 11 is a side view showing an example of the configuration of a blood vessel image capturing device according to a third embodiment. [Figure 8] FIG. 11 is a side view showing another example of the configuration of the blood vessel image capturing device of the third embodiment. [Figure 9] FIG. 11 is a side view showing another example of the configuration of the blood vessel image capturing device of the third embodiment. [Figure 10A] 13A to 13C are diagrams for explaining finger vascular pattern matching using the vascular image capturing device of the fourth embodiment without using additional joint position information. [Figure 10B] 13A to 13C are diagrams for explaining matching of finger vascular patterns using additional joint position information in the vascular image capturing device of the fourth embodiment. [Figure 11A] FIG. 13 is a diagram showing an example of a display screen when an outstretched finger is placed over the blood vessel image capturing device of the fifth embodiment. [Figure 11B] FIG. 13 is a diagram showing an example of a display screen when a bent finger is placed over the blood vessel image capturing device of the fifth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0012] Hereinafter, the embodiments will be described with reference to the drawings. EXAMPLES
[0013] This embodiment is an example of a vascular imaging device that captures the vascular pattern of a user's finger and has a function of detecting when the user is bending his / her finger (hereinafter referred to as "finger bending") and correcting the finger vascular pattern that has become distorted due to finger bending.
[0014] FIG. 1 is a side view showing an example of the configuration of a blood vessel image capturing apparatus according to a first embodiment.
[0015] The blood vessel imaging device 100 comprises a plurality of light sources 102 that irradiate near-infrared light 103 onto the dorsal side of a finger 111 to image the blood vessels of the finger 111, an upper cover 101 that stores the light sources 102, a support 104 that supports the upper cover 101, a fingertip-side finger rest 105 and a base-side finger rest 106 on which the finger 111 is placed when using the blood vessel imaging device 100, an imaging device 110 for capturing an image of the finger blood vessels from the ventral side of the finger 111, an optical filter 112 that prevents the intrusion of dirt and dust and transmits the near-infrared light 103 for imaging, and a control device 113 that controls the light sources 102 and the imaging device 110.
[0016] When a finger 111 is held over blood vessel image capturing device 100, near-infrared light 103 is irradiated from light source 102 onto the back of finger 111. Near-infrared light 103 is blocked by finger 111, but some of it passes through finger 111. Since hemoglobin contained in blood blocks near-infrared light more strongly than other parts of the body, the blood vessel pattern appears as dark lines in the image captured by imaging device 110.
[0017] Furthermore, the blood vessel image capturing device 100 may include a reflected light source 108 that irradiates near-infrared light 109 onto the ventral side of the finger 111. The near-infrared light 109 irradiated from the reflected light source 108 is reflected from the surface of the finger 111, or enters the inside of the finger 111, is scattered, and then exits again from the surface of the finger 111. This near-infrared light 109 is captured by the imaging device 110.
[0018] Next, an example in which a user holds a bent finger 111 over the blood vessel image capturing device 100, the back side of the finger 111 is illuminated by the light source 102, and an image is captured by the image capturing device 110 will be described with reference to FIG.
[0019] FIG. 2A is a side view showing an example of how a bent finger is photographed using the blood vessel imaging device 100 of Example 1, FIG. 2B is a diagram showing an example of an image captured when an amount of light 20% weaker than the amount of light at which the blood vessels in the finger are clearly visible is irradiated, FIG. 2C is a diagram showing an example of an image captured when an amount of light 20% stronger than the amount of light at which the blood vessels in the finger are clearly visible is irradiated, and FIG. 2D is a diagram showing an example of an image captured when an amount of light 20% stronger than the amount of light at which the blood vessels in the finger are clearly visible is irradiated.
[0020] 2A, near-infrared light 103 is irradiated from light source 102 to the back side of bent finger 111, and an image is captured by imaging device 110. Control device 113 adjusts the light intensity of light source 102 to an intensity 20% weaker than the intensity at which the blood vessels of finger 111 are clearly visible, an intensity at which the blood vessels of finger 111 are clearly visible, and an intensity 20% stronger than the intensity at which the blood vessels of finger 111 are clearly visible, and captures the images, thereby acquiring captured images 202, 206, and 210 shown in FIGS. 2B, 2C, and 2D.
[0021] 2B, 2C, and 2D, near-infrared light 103 easily passes through the first and second joints of bent finger 111, resulting in the appearance of narrow band-like areas (hereinafter referred to as "brightness saturation areas") 204, 205, 208, 209, 212, and 213 where brightness saturation has occurred, extending in the width direction of finger 111 at the joints. Brightness saturation refers to a situation where the brightness of an image captured by imaging device 110 exceeds the upper limit of brightness that imaging device 110 can represent, and the image is rounded to the upper limit that imaging device 110 can represent, even though it is actually brighter.
[0022] In each of the captured images 202, 206, and 210 in FIGS. 2B, 2C, and 2D, brightness saturated areas 204, 205, 208, 209, 212, and 213 appear clearly at the first and second joints of the bent finger 111.
[0023] When an outstretched finger 111 is held over the blood vessel image capturing device 100, a brightness saturation region appears spreading in both the vertical and horizontal directions, and when the light intensity of the light source 102 is increased, the brightness saturation region spreads in both the vertical and horizontal directions, and when the light intensity of the light source 102 is decreased, the brightness saturation region shrinks in both the vertical and horizontal directions.
[0024] On the other hand, when finger 111 is bent and held over blood vessel image capturing device 100, a band-like luminance saturated region appears clearly in the width direction of finger 111, which has the same width as finger 111 even when the light amount is changed within a range of ±20%, near the joint of finger 111. Therefore, control device 113 can detect that finger 111 is bent and held over blood vessel image capturing device 100 by detecting this luminance saturated region.
[0025] The band-shaped luminance saturated regions are characterized in that two appear parallel to each other at the first and second joints of the finger 111, separated by the same distance as the distance between the two joints. In addition, each of the two luminance saturated regions is long in the finger width direction, and is characterized in that it is perpendicular to the finger length direction at the length of the finger width. In addition, since the finger blood vessels run generally in the finger length direction to deliver blood to the finger's extremities, the band-shaped luminance saturated regions are also characterized in that they are perpendicular to the overall flow direction of the finger blood vessels.
[0026] These features allow the control device 113 to detect finger bending by identifying the brightness saturation regions 204, 205, 208, 209, 212, and 213 of the joints caused by finger bending.
[0027] Furthermore, the control device 113 can detect the positions of the first and second joints of the finger 111 by detecting the positions of the band-shaped luminance saturated regions 204, 205, 208, 209, 212, and 213 of the captured images 202, 206, and 210 simultaneously with detecting the bending of the finger.
[0028] In addition, when finger 111 is bent and held over blood vessel image capturing device 100, the tip and base of the finger appear darker in the image than when finger 111 is straightened and held over blood vessel image capturing device 100; and when finger 111 is bent and held over blood vessel image capturing device 100, the tip and base of the finger appear closer to image capturing device 110 and are enlarged, making the finger appear wider, and the finger appear narrower in the central part of the finger. By utilizing this fact in conjunction with detection of band-shaped brightness saturation regions 204, 205, 208, 209, 212, 213, the accuracy of detecting bent fingers can be improved.
[0029] The control device 113 also obtains the angles from the first joint to the tip of the finger 111 and from the second joint to the base of the finger to obtain the degree of finger bending, such as whether the finger is bent significantly or slightly. To obtain the degree of finger bending, the control device 113 checks the rate of change in finger width from the first joint toward the tip of the finger and from the second joint toward the base of the finger depending on the distance from the joint, and compares this with the average rate of change in finger width obtained in advance to calculate the angles from the first joint to the tip of the finger and from the second joint to the base of the finger.
[0030] In addition, the rate of change in luminance of finger 111 depending on the distance from the joint from the first joint toward the fingertip side and from the second joint toward the base of the finger may be examined, and the angle from the first joint toward the fingertip side and from the second joint toward the base of the finger may be calculated by comparing the rate of change in luminance with an average rate of change in luminance obtained in advance.
[0031] In addition, machine learning may be used to learn to estimate whether the finger is bent and the angles from the first joint to the tip of the finger and from the second joint to the base of the finger from finger 111 based on finger vein images of finger 111 bent at various angles and finger vein images of finger 111 extended, and the learning results may be used to estimate from a captured image whether the finger is bent and the angles from the first joint to the tip of the finger and from the second joint to the base of the finger.
[0032] Next, an example in which a user holds a bent finger 111 over blood vessel image capturing apparatus 100, and the belly side of finger 111 is illuminated by reflected light source 108 and captured by imaging device 110 will be described with reference to FIG.
[0033] FIG. 3A is a side view showing another example of how a bent finger is photographed using the blood vessel image photographing device 100 of Example 1, FIG. 3B is a diagram showing an example of a photographed image when an amount of light that does not cause any brightness saturation areas other than the reflection of the reflected light light source 108 is irradiated, FIG. 3C is a diagram showing an example of a photographed image when an amount of light 1.5 times the amount of light that does not cause any brightness saturation areas other than the reflection of the reflected light light source 108 is irradiated, and FIG. 3D is a diagram showing an example of a photographed image when an amount of light twice the amount of light that does not cause any brightness saturation areas other than the reflection of the reflected light light source 108 is irradiated.
[0034] 3A, near-infrared light 109 is irradiated from a reflected light source 108 onto the ventral side of a bent finger 111, and an image is captured by an imaging device 110. The control device 113 adjusts the light amount of the reflected light source 108 to an amount of light where no luminance saturated area other than the reflected light source 108 is generated, 1.5 times the amount of light where no luminance saturated area other than the reflected light source 108 is generated, and twice the amount of light where no luminance saturated area other than the reflected light source 108 is generated, and captures the images, thereby acquiring captured images 301, 308, and 316 shown in FIGS. 3B, 3C, and 3D.
[0035] 3B, 3C, and 3D, linear dark areas 306, 307, 313, 314, 321, and 322 caused by joint wrinkles appear at the first and second joints of bent finger 111. By detecting linear dark areas 306, 307, 313, 314, 321, and 322 caused by joint wrinkles, control device 113 can detect finger bending.
[0036] The linear dark areas 306, 307, 313, 314, 321, 322 due to joint wrinkles are characterized in that they appear in parallel at the first and second joints, separated by the same distance as the distance between the two joints, similar to the band-shaped brightness saturation areas 204, 205, 208, 209, 212, 213 in FIG. 2. In addition, each of the linear dark areas due to the two joint wrinkles is long in the finger width direction, and is perpendicular to the finger length direction at the length of the finger width. In addition, since the finger blood vessels run generally in the length direction of the finger 111 to deliver blood to the extremities of the finger 111, the linear dark areas 306, 307, 313, 314, 321, 322 due to joint wrinkles are characterized in that they are also perpendicular to the overall flow direction of the finger blood vessels.
[0037] The control device 113 can detect finger bending by identifying linear dark areas 306, 307, 313, 314, 321, and 322 caused by joint wrinkles due to finger bending using these features.
[0038] Furthermore, in the captured images 301, 308, and 316 of FIGS. 3B, 3C, and 3D, there appear central high brightness portions 303, 310, and 318 of the finger 111, as well as brightness saturated regions 304, 305, 311, 312, 315, 319, 320, and 323.
[0039] The circular brightness saturation regions 304 , 305 , 311 , 312 , 319 , and 320 are regions in which near-infrared light 109 emitted from the reflected light source 108 is reflected by the optical filter 112 and captured in the captured image, and regions in which near-infrared light 109 is reflected by the finger 111 overlap.
[0040] In the captured image 308 of FIG. 3C, in a high brightness area 310 in the center of the finger 111, in addition to the brightness saturation areas 311, 312 caused by the reflected light source 108 being reflected by the optical filter 112, a brightness saturation area 315 has occurred in the center of the finger 111.
[0041] In the captured image 316 of FIG. 3D, in a high-luminance portion 318 in the center of the finger 111, in addition to the brightness-saturated regions 319 and 320 caused by the reflected light source 108 being reflected by the optical filter 112, a brightness-saturated region 323 larger than the brightness-saturated region 315 of FIG. 3C has occurred in the center of the finger 111.
[0042] When an outstretched finger 111 is held over the blood vessel image capturing device 100, a brightness saturation region is generated due to the reflected light from the reflected light source 108 being reflected by the optical filter 112, and when the light intensity of the light source 102 is increased, a brightness saturation region is generated in the central part of the finger 111, and when the light intensity of the light source 102 is further increased, these brightness saturation regions become even larger, forming one large brightness saturation region.
[0043] On the other hand, when finger 111 is bent and held over blood vessel image capturing device 100, independent luminance saturated regions 319, 320, 323 appear even if the amount of light is doubled. Therefore, control device 113 may improve the accuracy of detecting finger bending by utilizing the fact that independent luminance saturated regions 319, 320, 323 appear even if the amount of light is increased, together with the detection of linear dark portions 306, 307, 313, 314, 321, 322 due to joint wrinkles.
[0044] Next, an example of correcting distortion of a finger vascular pattern caused by finger bending when finger bending is detected by the blood vessel image capturing device 100 will be described with reference to FIG.
[0045] FIG. 4A is a diagram showing an example of a finger vascular image captured by the vascular image capturing device 100 of Example 1 with the finger 111 bent, FIG. 4B is a diagram showing another example of a finger vascular image captured by the vascular image capturing device 100 of Example 1 with the finger 111 bent, and FIG. 4C is a diagram showing an example of a finger vascular image corrected by the vascular image capturing device 100 of Example 1.
[0046] Finger vascular images 401, 408 in Figs. 4A and 4B are finger vascular images captured by imaging device 110 when finger 111 is bent and held over vascular image capturing device 100, and the finger vascular pattern is captured as dark lines.
[0047] In the finger vascular image 401 of FIG. 4A, in the width direction of the finger 111, compared to when the finger 111 is extended and held over the vascular image capturing device 100, the central portion between the first and second joints of the finger 111 is farther away from the imaging device 110 and appears smaller, while the fingertip and base are relatively closer to the imaging device 110 and appear enlarged toward the fingertip and base.
[0048] In addition, in the finger vein image 401, in the longitudinal direction of the finger 111, the central part between the two joints is positioned higher in the vertical direction, and therefore the first joint to the fingertip side and the second joint to the base of the finger are oblique in the vertical direction and are therefore reduced in size.
[0049] 4A, in finger vein image 401 distorted by finger bending, control device 113 corrects the distortion of the finger vein pattern by stretching the image from the first joint to the fingertip side and from the second joint to the base side in the length direction of finger 111 in the directions of arrows 402 and 403, and shrinking the image in the width direction of finger 111 in the directions of arrows 404, 405, 406, and 407 so as to shrink the image toward the ends of the fingertip side and base side. This correction can be achieved by general image keystone correction.
[0050] 4B, depending on the shape of base-side finger rest 106 and the properties of finger 111, when finger 111 is stretched out and held over blood vessel image capturing device 100, finger 111 may be crushed at the base side, causing the base side to spread in the finger width direction, and when finger 111 is bent and held over blood vessel image capturing device 100, the base side of the finger may be tilted in the vertical direction and raised, causing the base side to be captured as being smaller in the finger width direction compared to when finger 111 is stretched out. Therefore, in finger blood vessel image 408, the base side appears thinner than the fingertip and the center of the finger.
[0051] In such a case, as shown in Fig. 4B, in finger vein image 408 distorted by finger bending, control device 113 stretches finger 111 lengthwise from the first joint to the fingertip side and from the second joint to the base side in the directions of arrows 409 and 410 as in Fig. 4A, shrinks finger 111 widthwise in the directions of arrows 412 and 414 so as to shrink the fingertip side end, and stretches finger 111 widthwise in the directions of arrows 411 and 413 so as to expand the finger base side end, thereby correcting the distortion of the finger vein pattern. This correction can be achieved by using the same keystone correction as in Fig. 4A, but changing the direction of keystone correction on the base side from that in Fig. 4A.
[0052] Distortion correction may be performed by learning in advance through machine learning to estimate an undistorted finger vascular image with an extended finger from a finger vascular image with distortion caused by bending a finger using a finger vascular image with distortion caused by bending a finger and a finger vascular image with no distortion caused by straightening the finger, and then using the learning results to estimate an undistorted finger vascular image with an extended finger from a finger vascular image with distortion caused by bending a finger.
[0053] According to this embodiment, the control device 113 detects finger bending by detecting brightness saturated areas at the joints caused by bending the finger, or linear dark areas caused by joint wrinkles, and corrects the vascular image by stretching the fingertip and base sides of the finger vascular image in the finger length direction, stretching the fingertip side of the finger vascular image in the finger width direction, and stretching or reducing the base side of the vascular image in the finger width direction, thereby making it possible to properly authenticate the user even when the user holds a bent finger over the vascular image capturing device. EXAMPLES
[0054] This embodiment is an example of an individual authentication device system using the blood vessel image capturing device 100 of the first embodiment.
[0055] Fig. 5 is a diagram showing an example of the configuration of an individual authentication system according to Example 2. In Fig. 5, the same components as those in Fig. 1 are given the same reference numerals, and the description thereof will be omitted.
[0056] 5 includes a blood vessel image capturing device 100 and an authentication device 500 connected to the blood vessel image capturing device 100. In the first embodiment, the control device 113 of the blood vessel image capturing device 100 controls the light source 102 and detects finger bending, but in the second embodiment, these processes are executed by the authentication device 500.
[0057] The authentication device 500 includes a near-infrared light image input unit 502, a near-infrared light source control unit 503, an interface 504, an external storage device 505, a CPU (Central Processing Unit) 506, a memory 507, a display device 508, a keyboard 509, and a speaker 510.
[0058] When a user holds a finger 111 over the blood vessel image capturing device 100 connected to the authentication device 500, the image captured by the imaging device 110 through the optical filter 112 is converted into an electrical signal, which is then input as an image to the CPU 506 via the near-infrared light image input unit 502 and the interface 504 and recorded in the memory 507.
[0059] By executing a program stored in memory 507, CPU 506 detects from the captured image recorded in memory 507 that weak light emitted from light source 102 has been blocked by finger 111, and identifies that finger 111 has been held over blood vessel image capturing device 100.
[0060] When it is recognized that a finger 111 is held up, the CPU 506 executes a program stored in the memory 507 to control the near-infrared light source control unit 503 via the interface 504 to irradiate near-infrared light 103 sufficient to capture an image of the finger blood vessels from the light source 102.
[0061] An image of a finger 111 irradiated with near-infrared light 103 from a light source 102 passes through an optical filter 112 and is converted into an electrical signal by an imaging device 110. The image is then taken into a CPU 506 via a near-infrared light image input unit 502 and an interface 504, and is recorded in a memory 507.
[0062] CPU 506 executes a program recorded in memory 507 to compare the image recorded in memory 507 with one or more finger vein patterns or the like registered in advance in external storage device 505, and performs identity authentication. The result of identity authentication is notified to the user by voice from speaker 510, or by displaying it on display device 508 in response to an instruction from keyboard 509.
[0063] In this embodiment, the authentication device 500 and the vascular image capturing device 100 are described as separate devices, but they may be configured as a vascular image capturing device including an image capturing unit that captures a finger vascular image and an authentication unit that performs identity authentication based on information from the captured finger vascular image.
[0064] 6 is a flowchart showing an example of personal authentication processing executed by the personal authentication device of Example 2. This processing is realized by the CPU 506 executing a program stored in the memory 507.
[0065] First, when a user holds finger 111 over blood vessel image capturing device 100, near-infrared light 103 is irradiated from light source 102 onto the dorsal side of finger 111 (step S602), and the imaging device 110 captures an image from the ventral side of finger 111 to obtain a captured image (step S603).
[0066] Next, a finger is detected from the captured image acquired in step S603 (step S604), and it is determined whether or not a finger is detected from the captured image in step S604 (step S605). If it is determined in step S605 that a finger is detected from the captured image, the process proceeds to step S606, and if it is determined in step S605 that a finger is not detected from the captured image, the process returns to step S602.
[0067] In step S606, the average luminance of the detected finger is calculated in the acquired captured image. For example, when a finger is detected in the captured image, the average luminance of the finger is calculated by calculating the average luminance of a portion of the captured image where the finger is always captured.
[0068] Next, it is determined whether the average brightness of the finger calculated in step S606 is between the upper and lower limits of the target brightness (step S607). If the average brightness of the finger is not between the upper and lower limits of the target brightness in step S607, the process proceeds to step S608, and if the average brightness of the finger is between the upper and lower limits of the target brightness in step S607, the process proceeds to step S609.
[0069] In step S608, the light source 102 is controlled so as to increase the light intensity value if the average luminance of the finger is below the lower limit of the target luminance, and to reduce the light intensity value if the average luminance of the finger is above the upper limit of the target luminance, and then the process returns to step S602.
[0070] 2 or 3 from the captured image, it is detected whether the user is bending his / her finger 111 and holding it over the blood vessel image capturing device 100, and the degree of finger bending is detected if the user is bending his / her finger 111 and holding it over the blood vessel image capturing device 100. The degree of finger bending is detected by calculating the angle from the first joint to the tip of the finger and the angle from the second joint to the base of the finger, as described in FIG.
[0071] Next, based on the degree of finger bending detected in step S609, distortion of the finger blood vessel image that occurs when the user bends his or her finger and holds it over the blood vessel image capturing device 100 is corrected (step S610).
[0072] Next, the finger contour is detected from the finger blood vessel image to clarify the area of the finger 111 in the finger blood vessel image (step S611), and a finger blood vessel pattern for matching is generated from the finger blood vessel image (step S612).
[0073] Next, the finger vein pattern 614 registered in advance in the external storage device 505 is compared with the finger vein pattern generated from the captured image in step S612 (step S613), and the score of the degree of mismatch between the vein patterns is calculated as the matching score. The matching score is calculated, for example, by performing template matching to check whether a pattern of one image exists in another image, checking the number of pixels that match the registered finger vein pattern 614 while gradually shifting the position of the finger vein pattern to be checked, and counting the number of mismatching pixels at the position with the most matching pixels.
[0074] Next, it is determined whether the calculated matching score is below the threshold value (step S615). If it is determined in step S615 that the calculated matching score is below the threshold value, it is determined that the generated finger vein pattern matches the registered finger vein pattern 614, and the process proceeds to step S616, where post-successful authentication processing is performed, and the process ends.
[0075] On the other hand, if it is determined in step S615 that the calculated matching score is not below the threshold, it is determined that the generated finger vein pattern does not match the registered finger vein pattern 614, and the process proceeds to step S617, where post-authentication failure processing is performed, and the process ends.
[0076] According to this embodiment, by detecting brightness saturated areas or dark parts from the captured image, it is detected whether the user is bending his / her finger 111 and holding it over the blood vessel image capturing device 100. If the user bends his / her finger 111 and holds it over the blood vessel image capturing device 100, the degree of finger bending is detected (step S609). The distortion of the finger blood vessel image is corrected from the detected degree of finger bending (step S610). Therefore, the user can be properly authenticated even if he / she bends his / her finger and holds it over the blood vessel image capturing device. EXAMPLES
[0077] This embodiment is an example in which the arrangement of the light source is changed in the blood vessel image capturing device of embodiment 1 in order to capture a clear finger blood vessel image while improving the accuracy of detecting finger bending and the degree of finger bending.
[0078] Fig. 7 is a side view showing an example of the configuration of a blood vessel image capturing apparatus 100 of embodiment 3. In Fig. 7, the same components as those in Fig. 1 are given the same reference numerals, and the description thereof will be omitted.
[0079] 7 further includes a fingertip side light source 701 and a base side light source 702 in addition to the configuration in Fig. 1. Fingertip side light source 701 is arranged tilted toward the fingertip from a vertically downward direction, and irradiates near-infrared light 703 onto the dorsal side of finger 111. Base side light source 702 is arranged tilted toward the base side from a vertically downward direction, and irradiates near-infrared light 704 onto the dorsal side of finger 111.
[0080] When a finger 111 is bent and held over the blood vessel image capturing device 100, near-infrared light 703, 704 is irradiated onto the fingertip and base of the finger at an angle onto the surface of the finger 111. This results in a longer path for the near-infrared light 703, 704 to pass through the inside of the finger 111, and the fingertip and base of the finger tend to appear dark in the image captured by the imaging device 110.
[0081] This increases the difference in brightness between the fingertip and base when finger 111 is extended and held over blood vessel image capturing device 100 and when finger 111 is bent and held over blood vessel image capturing device 100, thereby improving the accuracy of detecting finger bending and the degree of finger bending.
[0082] The fingertip side light source 701 is preferably arranged so that the optical axis angle 705 from the vertical direction is in the range of 10 to 80 degrees on the fingertip side. The finger base side light source 702 is preferably arranged so that the optical axis angle 706 from the vertical direction is in the range of 10 to 80 degrees on the finger base side.
[0083] In the example of FIG. 7, a fingertip-side light source 701 and a base-side light source 702 are further provided, so that the difference in brightness between the fingertip side and the base side becomes greater when finger 111 is straightened and held over blood vessel image capturing device 100 compared to when finger 111 is bent and held over blood vessel image capturing device 100. This improves the accuracy of detecting finger bending and the degree of finger bending, and enables correct identity authentication even when a user bends his or her finger and holds it over the blood vessel image capturing device.
[0084] Fig. 8 is a side view showing another example of the configuration of the blood vessel image capturing apparatus 100 of the embodiment 3. In Fig. 8, the same components as those in Fig. 1 are given the same reference numerals, and the description thereof will be omitted.
[0085] 8 further includes a fingertip side light source 801 and a base side light source 802 in addition to the configuration in Fig. 1. Fingertip side light source 801 is arranged tilted toward the base of the finger from a vertically downward direction, and irradiates near-infrared light 804 onto the dorsal side of finger 111. Furthermore, base side light source 802 is arranged tilted toward the fingertip from a vertically downward direction, and irradiates near-infrared light 803 onto the dorsal side of finger 111.
[0086] 7, the inclination from the vertical of the fingertip-side light source 701 and the base-side light source 702 is greater. Therefore, even if the finger 111 is bent only slightly, the near-infrared light 803, 804 irradiated to the fingertip and base is irradiated at an angle to the surface of the finger, so that the path taken by the near-infrared light 803, 804 inside the finger 111 becomes longer, and the fingertip and base tend to appear dark in the image captured by the imaging device 110.
[0087] This increases the difference in brightness between the fingertip and base when finger 111 is extended and held over blood vessel image capturing device 100 and when finger 111 is bent and held over blood vessel image capturing device 100, thereby improving the accuracy of detecting finger bending and the degree of finger bending.
[0088] Fingertip side light source 801 is preferably arranged so that optical axis angle 805 from the vertical direction is in the range of 10 to 80 degrees toward the base of the finger. Also, finger base side light source 802 is preferably arranged so that optical axis angle 706 from the vertical direction is in the range of 10 to 80 degrees toward the fingertip.
[0089] According to the example of FIG. 8, a fingertip-side light source 801 and a base-side light source 802 are further provided, so that even if the finger 111 is only slightly bent, the difference in brightness between the fingertip side and the base side becomes large when the finger 111 is straightened and held over the blood vessel image capturing device 100 and when the finger 111 is bent and held over the blood vessel image capturing device 100. This improves the accuracy of detecting finger bending and the degree of finger bending, and enables correct identity authentication even when the user bends his or her finger and holds it over the blood vessel image capturing device.
[0090] Fig. 9 is a side view showing another example of the configuration of the blood vessel image capturing apparatus 100 of the embodiment 3. In Fig. 9, the same components as those in Fig. 1 are given the same reference numerals, and the description thereof will be omitted.
[0091] 9 further includes a fingertip-side additional light source 901 and a base-side additional light source 902 in addition to the configuration in Fig. 1. The fingertip-side additional light source 901 is disposed tilted toward the base of the finger from a vertically downward direction, and irradiates near-infrared light 903 onto the back side of the finger 111. The base-side additional light source 902 is disposed tilted toward the fingertip from a vertically downward direction, and irradiates near-infrared light 904 onto the back side of the finger 111.
[0092] This prevents a situation in which, when a user holds a bent finger 111 over the blood vessel image capturing device 100, the fingertip and base of the finger become dark, making the finger vessel image unclear and making it impossible to extract an accurate finger vessel pattern.
[0093] The fingertip side additional light source 901 is preferably arranged so that the optical axis angle 905 from the vertical direction is in the range of 10 to 80 degrees toward the base of the finger. The fingertip side additional light source 902 is preferably arranged so that the optical axis angle 906 from the vertical direction is in the range of 10 to 80 degrees toward the fingertip.
[0094] According to the example of FIG. 9, by further providing an additional light source 901 on the fingertip side and an additional light source 902 on the base side, when a user bends his / her finger 111 and holds it over the blood vessel image capturing device 100, it is possible to prevent the fingertip and base from becoming dark, which would result in an unclear finger vessel image and an inability to extract an accurate finger vessel pattern. Therefore, it is possible to properly authenticate the user even when he / she holds his / her bent finger over the blood vessel image capturing device. EXAMPLES
[0095] This embodiment is an example in which the search range for aligning finger vein patterns is narrowed by using the positions of the first and second joints of the finger, which are acquired when detecting finger bending, thereby shortening the processing time required to match finger vein patterns.
[0096] FIG. 10A is a diagram for explaining matching of finger vascular patterns without using joint position additional information in the vascular image capturing device of the fourth embodiment, and FIG. 10B is a diagram for explaining matching of finger vascular patterns using joint position additional information in the vascular image capturing device of the fourth embodiment.
[0097] In FIG. 10A, a registered blood vessel pattern 1001 registered in the external storage device 505 is compared with an input blood vessel pattern 1002 generated from a captured image by the blood vessel image capturing device 100.
[0098] The matching between the registered vascular pattern 1001 and the input vascular pattern 1002 is performed, for example, by performing template matching to check whether a certain image pattern exists in another image, and by shifting the position of the image pattern to be checked slightly forward, backward, left, and right to check the number of pixels that match the other image, and by calculating the matching score by counting the number of non-matching pixels in the position with the most matching pixels, and then determining the lowest matching score to perform the matching.
[0099] A search range 1003 is a range in which the position of the input blood vessel pattern 1002 is superimposed on the registered blood vessel pattern 1001 while being shifted slightly back and forth and left and right to search for the position with the smallest matching score.
[0100] Next, in FIG. 10B, a registered blood vessel pattern 1004 registered in the external storage device 505 is collated with an input blood vessel pattern 1007 generated from a captured image by the blood vessel image capturing device 100.
[0101] 2, when detecting finger bending, the positions of the first joint and the second joint of the finger 111 can be detected by detecting the positions of the band-shaped brightness saturation region of the captured image. In the example of FIG. 10B, information indicating the positions of the first joint and the second joint is added to the finger vein pattern as first joint position additional information and second joint position additional information.
[0102] A registered blood vessel pattern 1004 is added with first joint position additional information indicating a first joint position 1005 and second joint position additional information indicating a second joint position 1006. Moreover, an input blood vessel pattern 1007 is added with first joint position additional information indicating a first joint position 1008 and second joint position additional information indicating a second joint position 1009.
[0103] When matching the registered vascular pattern 1004 with the input vascular pattern 1007 by template matching, first, the longitudinal positions of the finger 111 of the registered vascular pattern 1004 and the input vascular pattern 1007 are aligned so that the first joint positions 1005, 1008 and the second joint positions 1006, 1009 match.
[0104] Specifically, the registered blood vessel pattern 1004 and the input blood vessel pattern 1007 are aligned in the longitudinal direction of the finger 111 so that a first joint position 1005 and a first joint position 1008 and a second joint position 1006 and a second joint position 1009 match.
[0105] This eliminates the need to align the registered vein pattern 1004 and the input vein pattern 1007 in the length direction of the finger 111, and only aligns the width direction of the finger 111. Therefore, the search range 1010 is smaller than the search range 1003 in Fig. 10A, and the processing time for personal authentication can be shortened.
[0106] According to this embodiment, the longitudinal positions of the finger 111 of the registered vein pattern 1004 and the input vein pattern 1007 are aligned so that the first joint positions 1005, 1008 and the second joint positions 1006, 1009 match, so that the search range 1010 is reduced and the processing time for the personal authentication process can be shortened. EXAMPLES
[0107] This embodiment is an example of a finger vein imaging device that, when it detects that a finger is bent, notifies the user that the finger is bent and prompts the user to straighten the finger, thereby enabling proper identity authentication.
[0108] Fig. 11A is a diagram showing an example of a display screen when an outstretched finger is held over the blood vessel image capturing device 100 of the embodiment 5, and Fig. 11B is a diagram showing an example of a display screen when a bent finger is held over the blood vessel image capturing device 100 of the embodiment 5. In Fig. 11A and Fig. 11B, the same components as those in Fig. 1 are given the same reference numerals, and the description thereof will be omitted.
[0109] The blood vessel image capturing device 100 in FIGS. 11A and 11B further includes a display device 1101 in addition to the configuration in FIG.
[0110] 11A, when a user holds outstretched finger 111 over blood vessel imaging device 100, control device 113 detects that finger 111 is being held over blood vessel imaging device 100, and causes display device 1101 to illuminate display 1102 informing that finger 111 has been held over blood vessel imaging device 100 and to darken display 1103 informing that finger 111 has been bent and held over blood vessel imaging device 100. This notifies the user that finger 111 has been held over blood vessel imaging device 100.
[0111] 11B, when the user bends and holds finger 111 over blood vessel imaging device 100, control device 113 detects that finger 111 has been bent and held over blood vessel imaging device 100, and darkens display 1102 on display device 1101, informing that finger 111 has been straightened and held over blood vessel imaging device 100, and brightens display 1103, informing that finger 111 has been bent and held over blood vessel imaging device 100. This makes it possible to inform the user that finger 111 has been bent and held over blood vessel imaging device 100, and to prompt the user to straighten finger 111.
[0112] In addition, in FIG. 11, an example is described in which the fact that finger 111 is extended and held over blood vessel imaging device 100, or that finger 111 is bent and held over blood vessel imaging device 100, is displayed on display device 1101 provided in blood vessel imaging device 100 under the control of control device 113, but it may also be displayed on display device 508 provided in authentication device 500 in FIG. 5 under the control of CPU 505.
[0113] According to this embodiment, the user is informed that he / she has bent his / her finger 111 and is prompted to straighten his / her finger 111 over the blood vessel image capturing device 100, so that the user can be properly authenticated by straightening his / her finger 111 and re-holding it over the blood vessel image capturing device 100.
[0114] The present invention is not limited to the above-described embodiments, and includes various modified examples. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the configurations described. It is also possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration. [Explanation of symbols]
[0115] 100 Vascular imaging device 102 Light source 103 Near infrared light 110 Imaging device 111 fingers 113 Control device
Claims
1. A blood vessel imaging device for capturing an image of blood vessels in a user's finger, A light source that irradiates light onto the back side of the finger; an imaging means for imaging the finger irradiated with the light from the pad side of the finger; and a control unit that detects whether the finger is bent by detecting an area where a band-shaped brightness saturation has occurred that has passed through the finger joint from the captured blood vessel image.
2. 2. The blood vessel imaging apparatus according to claim 1, The finger touch sensor further includes a reflected light source that irradiates light onto the pad side of the finger, The imaging means captures an image of the finger irradiated with light from the reflected light source from a pad side of the finger, The control unit detects that the finger is bent by detecting linear dark portions due to wrinkles at the finger joints from the captured blood vessel image.
3. 2. The blood vessel imaging apparatus according to claim 1, The control unit is When it is detected that the finger is bent, the blood vessel image is corrected by stretching the fingertip side and base side of the blood vessel image in the finger length direction, stretching the fingertip side of the blood vessel image in the finger width direction, and stretching or shrinking the base side of the blood vessel image in the finger width direction.
4. 2. The blood vessel imaging apparatus according to claim 1, The light source includes a fingertip-side light source provided on the fingertip side and a finger base-side light source provided on the finger base side, the fingertip-side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees toward the fingertip; The finger base side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees on the finger base side.
5. 2. The blood vessel imaging apparatus according to claim 1, The light source includes a fingertip-side light source provided on the fingertip side and a finger base-side light source provided on the finger base side, the fingertip-side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees toward the base of the finger, The finger base side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees toward the fingertip side.
6. 2. The blood vessel imaging apparatus according to claim 1, The control unit is The blood vessel imaging device detects the position of the finger joint by detecting the position of a band-like area of luminance saturation that has passed through the finger joint from the captured blood vessel image.
7. 2. The blood vessel imaging apparatus according to claim 1, The blood vessel image capturing apparatus further comprises a display unit that notifies the user that the control unit has detected that the finger is bent.
8. a blood vessel image capturing device for capturing an image of blood vessels in a user's finger; An authentication system including an authentication device that performs authentication using a blood vessel image captured by the blood vessel image capturing device, The blood vessel image capturing device comprises: A light source that irradiates light onto the back side of the finger; an imaging means for imaging the finger irradiated with the light from the pad side of the finger; a control unit that detects whether the finger is bent by detecting a region in which a band-like luminance saturation occurs through the finger joint portion from the captured blood vessel image, The authentication device includes: A storage unit that stores a registered vascular pattern that has been registered in advance; a processing unit that generates a finger vein pattern from the captured vein image and performs identity authentication by comparing the generated finger vein pattern with the registered vein pattern.
9. 9. The personal authentication system according to claim 8, The blood vessel image capturing device further includes a reflected light source that irradiates light onto the pad side of the finger, The imaging means captures an image of the finger irradiated with light from the reflected light source from a pad side of the finger, The control unit detects whether the finger is bent by detecting linear dark areas caused by wrinkles at the finger joints from the captured blood vessel image.
10. 9. The personal authentication system according to claim 8, The control unit is When it is detected that the finger is bent, the personal authentication system corrects the blood vessel image by stretching the fingertip and base sides of the blood vessel image in the finger length direction, stretching the fingertip side of the blood vessel image in the finger width direction, and stretching or shrinking the base side of the blood vessel image in the finger width direction.
11. 9. The personal authentication system according to claim 8, The light source includes a fingertip-side light source provided on the fingertip side and a finger base-side light source provided on the finger base side, the fingertip-side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees toward the fingertip; The finger base side light source is disposed so that the optical axis angle from the vertical direction is within a range of 10 to 80 degrees on the finger base side.
12. 9. The personal authentication system according to claim 8, The light source includes a fingertip-side light source provided on the fingertip side and a finger base-side light source provided on the finger base side, the fingertip-side light source is disposed so that the optical axis angle from the vertical direction is in the range of 10 to 80 degrees toward the base of the finger, The finger base side light source is arranged so that the optical axis angle from the vertical direction is within a range of 10 to 80 degrees toward the fingertip side.
13. 9. The personal authentication system according to claim 8, The control unit is detecting a position of a region where a band-like luminance saturation occurs through the finger joint from the captured blood vessel image, thereby detecting the finger joint position; The processing unit includes: An identity authentication system that matches the generated finger vein pattern with the registered vein pattern by aligning the positions of the generated finger vein pattern and the registered vein pattern using the detected joint positions.
14. 9. The personal authentication system according to claim 8, The authentication device includes: The personal authentication system further comprises a display unit that notifies the user that the control unit has detected that the finger is bent.
Citation Information
Patent Citations
Ac generator for vehicle
JP1985005750A
