JP2006158952A - Biological feature input device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a small biological feature input device low in cost and capable of stably reading biological features such as fingerprints even in unfavorable conditions such as the dry or wet state of the skin or the peeled state of the skin. <P>SOLUTION: A one-dimensional or semi-one-dimensional image sensor 5 is used as an image sensor. A finger travel guide 3 is provided for keeping a nearly constant distance between a finger 4 and an effective picture element part 1 of the image sensor 5 without contacting each other during the relative sliding motion of the finger 4 and image sensor 5. One-dimensional or semi-one-dimensional partial images obtained by picking up images formed by radiation scattered inside the finger 4 and radiated from the skin surface of the finger 4, by the image sensor 5 during the relative motion, are connected by image processing in a microprocessor section 8 to reconstruct the image of the whole finger. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an apparatus for inputting a biometric feature for authenticating an individual, and more particularly to an apparatus for inputting an image of a biometric feature such as a fingerprint of a finger or other skin pattern or blood vessel pattern.

  Conventional biometric feature input devices for authenticating individuals using this type of finger are typically devices that read fingerprints, which are skin patterns on the fingertips, and are physical quantities such as light, electric field, pressure, capacitance, and temperature. Various readers using the absolute value and the amount of change have been developed.

  A method using a total reflection critical angle using a fiber optic plate (see, for example, Patent Document 2) or a prism (see, for example, Patent Document 3) is widely used as a fingerprint input device. With reference to FIG. 23, for example, an example using the total reflection critical angle of a conventional prism will be described. A lens 106 and a two-dimensional image sensor 107 are arranged in a direction in which light from the prism surface 108 is totally reflected by the prism surface 109 in the air with respect to the prism 105. The finger skin 104 is written with the skin pattern enlarged. The light 101 from the portion not in contact with the prism is largely refracted from the air having a refractive index of 1.0 and is incident on the prism surface 108 having a refractive index of 1.4 or more, and is totally reflected by the prism surface 109 or the prism 101. The surface 102 does not reach the two-dimensional image sensor 107, but the light 102 in contact with the skin is reflected on the prism surface 108 because the refractive index of oil or moisture on the skin or skin surface is close to that of the prism glass. The refraction angle becomes smaller and the prism surface 109 does not reach the total reflection angle. The image is formed by the lens 106 and reaches the two-dimensional image sensor 107. In this way, the shadow obtained by not touching the prism of the skin pattern such as the fingerprint of the finger is detected.

  Similarly, a fingerprint image is obtained by touching the skin, and a pressure, temperature, or capacitance type quasi-one-dimensional sensor is used to realize miniaturization, and a finger touching the quasi-one-dimensional sensor is moved. There has been proposed a technique for reconstructing a fingerprint image by connecting partial images of fingerprints of the fingers (see, for example, Patent Documents 1 and 7). In particular, methods using capacitance and temperature have already been put into practical use, and these methods contribute to downsizing and cost reduction of the device.

  Under such circumstances, a non-contact fingerprint detection apparatus has been proposed (see, for example, Patent Document 6). This is the same as a fingerprint, in which the incident light is scattered inside the finger and the radiation emitted from the finger again reflects the internal structure of the skin, with the concave portion of the fingerprint being the bright region and the convex portion being the dark region. This is based on the phenomenon that a gray image of a shape is obtained. According to this non-contact method, it is difficult to read the peeled portion in the method based on the above contact, and it is difficult to read even the finger peeled due to dermatitis etc. If the part of the structure is saved, the video can be obtained. Moreover, since it is non-contact, it is hard to receive the influence of the skin surface state changes, such as wetness and drying.

  In addition, the scattered radiation from the finger is imaged by a two-dimensional image sensor placed close to the finger through a transparent protective cover made of glass or the like, and the fingerprint concave portion is a dark portion region and the convex portion is a bright portion region. A fingerprint input device for acquiring an image has been proposed by the present inventor (see, for example, Patent Document 4). This is less susceptible to external environments such as finger wetting and drying and ambient light compared to sensors using pressure, temperature, capacitance, and total reflection critical angle. Further, as described in Patent Document 5 proposed by the present inventor, an image with high contrast can be obtained by optimally selecting the refractive index of the transparent protective cover.

On the other hand, in recent years, a technique for authenticating a blood vessel pattern on the base side of a finger below the first joint of the finger other than the fingerprint as an input device for biometric features existing in the finger has been put into practical use. This is a technique for reading thick blood vessel patterns such as veins using near-infrared absorption by blood. Optical CT (computer tomography), which was actively studied in the 1980s, is a computer of the human body by light that is less harmful to the human body. This is an application of technology to perform tomography. A blood vessel image is obtained by irradiating near infrared light from the upper part of the finger, and the light passing through the finger and emitted from the belly of the opposite finger darkens due to absorption of near infrared light in the blood abundantly present in the blood vessel. . This technique is an effective biometric authentication means when the fingerprint is not a problem, or the skin is peeled off due to dermatitis or the like, and the biometric feature input device based on contact is deteriorated and difficult to image.
JP-A-10-91769 Japanese Patent No. 3045629 US Pat. No. 6,381,347 Japanese Patent No. 3150126 JP 2003-006627 A JP 2003-85538 A JP 2001-155137 A

  In recent years, with the development of computerization, the leakage of personal information and the impersonation of others in transaction activities on the network have become problems. In order to prevent this, an apparatus for authenticating an individual by inputting a biometric characteristic unique to the individual has been developed instead of a method of easily allowing another person to impersonate by snooping or theft such as a password or an authentication card. In addition, downsizing and cost reduction of information handling devices such as mobile phones are progressing, and devices for inputting biometric features are also required to be downsized and reduced in price. In addition, considering application cases where biometric personal authentication is important for payment operations such as credit cards, the biometric feature input device can be made more accurate in order to authenticate the person reliably and under any circumstances. The need for is also increasing.

  Under such circumstances, fingerprints have not been the same for a long time, and the characteristics of lifelessness that will not change throughout the lifetime have been researched and confirmed, especially in the field of police and justice, and high-precision personal authentication is possible. However, the conventional fingerprint input device is difficult to obtain a good fingerprint image under bad conditions such as wet and dry fingers and peeling of skin due to dermatitis, etc., and it is difficult to say that it can be used by everyone. It was.

  The fingerprint input method using the total reflection critical angle by a fiber optic plate (for example, Patent Document 2) or a prism (for example, Patent Document 3) is widely used for personal authentication. Is obtained by contact between the unevenness of the skin and the prism, and the image of the peeled portion is lost. Moreover, expensive and large optical components are used, which hinders downsizing and cost reduction of the apparatus.

  There are several practical examples of pressure, electric field, or capacitance type two-dimensional sensors. Although optical components have been abolished and contributed to miniaturization and cost reduction, both images are premised on contact and the image of the peeled part is lost. In addition, there is a problem that it is weak against changes in the situation such as wetness and dryness of the finger compared to the optical method.

  Using a pressure, temperature, electric field, or capacitance type quasi-one-dimensional sensor, and sliding a finger in contact with the sensor to reconstruct the image of the fingerprint of the finger (for example, Patent Document 1 and Patent Document 7), Although it contributes to further downsizing and cost reduction of the device, the image of the part that is not in contact is lost, so if the skin is partially peeled off due to dermatitis, fingerprint authentication, that is, authentication by biometric features There was a problem that it was difficult. In addition, a method for reconstructing an image by moving a reading target using a one-dimensional sensor is known in a facsimile or a copying machine, but a special method for obtaining a speed in a direction of moving a finger to reduce the size of the apparatus. In order to omit an unnecessary mechanism, in Patent Document 1, a video is reconstructed by judging the similarity of images of several quasi-one-dimensional lines. An example of fingerprint image reconstruction in this method will be described with reference to FIG.

  A partial image from I1 to In is obtained by moving the finger with respect to the image of the finger 301, and a reconstructed image 302 is obtained by removing the similar portion. However, in this method, when the finger is moved slowly with respect to the imaging speed of the sensor as in the example shown in FIG. 25, the degree of overlap becomes large and it becomes difficult to judge the change in similarity. There is a problem that it is distorted by being extended. In addition, if the finger is slid faster than the imaging speed, an image missing between I1 and In as shown in FIG. 26 is generated, and the image may be shrunk vertically and distorted like the fingerprint image 304. There is a problem.

  A non-contact fingerprint detection apparatus has been proposed as an invention for improving the decrease in authentication accuracy due to the skin peeling (Patent Document 6). According to this, since the radiated light incident on the finger, scattered inside the finger and radiated from the skin surface of the finger reflects the internal structure of the skin, the same shade as the fingerprint is observed. This does not depend on the moist and dryness of the epidermis, and even when the epidermal stratum corneum has fallen off due to dermatitis etc., the fingerprint image can be obtained if the structure of the dermis that is the origin of the epidermal pattern such as fingerprints is preserved. It is done. However, in the case of the fingerprint detection apparatus disclosed in Patent Document 6, a target image cannot be obtained unless a space is provided between the finger and the imaging system so as to be non-contact. In addition, a finger fixing frame is necessary because of the necessity of focusing, which hinders operability and downsizing of the apparatus. In addition, an imaging optical system is required, and there is another problem that the apparatus becomes larger. In addition, the finger and the imaging system are separated from each other, and even if the amount of light emitted from the skin surface changes due to the internal structure of the finger, it is scattered on the skin surface. Therefore, there is a problem that a fingerprint image having a good contrast cannot be obtained at the part where the skin is actually directed.

  The fingerprint authentication device (Patent Document 4) made by the present inventor obtains a fingerprint image by imaging a radiation light scattered from a finger and emitted from the skin surface with a two-dimensional image sensor close to the finger. Therefore, downsizing and cost reduction of the device are realized. However, although a large two-dimensional image sensor is required and an optical system such as a lens has been eliminated, it has hindered further downsizing and cost reduction of the apparatus.

  Moreover, it is made clear by Patent Document 5 according to the proposal of the present inventor that the fingerprint image by the scattered radiation from the finger largely depends on the interface state between the skin and the sensor protective film. On the other hand, it is clear that the technique of reading scattered radiation from a finger reflects the structure inside the finger because the light is once incident inside the finger. Therefore, the fingerprint input device according to Patent Document 4 by the present inventor has been pointed out in Patent Document 6 as a phenomenon in a non-contact portion where the optical imaging system is eliminated and a small-sized fingerprint detection device is peeled off. An image reflecting the internal structure of the finger skin is obtained. However, the refractive index of the transparent cover existing between the fingerprint and the two-dimensional image sensor arranged in the vicinity thereof is set to be a light intensity corresponding to the convex portion of the fingerprint in contact with the transparent cover as described in Patent Document 5. If the contrast with the dark part area corresponding to the concave part not in contact with the part area is selected, the influence of reflection and refraction at the interface becomes strong and the component reflecting the skin structure becomes small. There is a problem that it is difficult to obtain the contrast of the fingerprint image reflecting the skin structure that originally appears. This problem is particularly noticeable when the dynamic range cannot be widened. If the non-contact state is maintained, the influence of the interface is eliminated. However, in a two-dimensional image sensor, it is difficult to obtain a stable fingerprint image because the non-contact state cannot be maintained at a certain distance with respect to a finger with curvature. there were.

  On the other hand, the technology that authenticates the blood vessel pattern on the base side of the finger below the first joint other than the fingerprint as an input device for the biometric features existing on the finger is not a problem with the fingerprint or the fingerprint due to dermatitis, etc. This is provided as an effective biometric authentication means when the image is deteriorated and difficult to be imaged. However, the effective amount of information of the pattern is generally smaller than various fingerprints, and changes depending on a nutrient state, a thrombus, a blood pressure or the like. Compared to fingerprints that have been researched mainly by the police and judiciary fields as a universal and lifelong, the accuracy is unconfirmed and remains as a future research subject. However, for example, as disclosed in Patent Document 7, if it can be read at the same time as a fingerprint pattern, it can be complemented with fingerprint information, or it can be an effective source of information as to whether or not it is a living body, and is also effective as a method for determining a fake finger. . However, like the non-contact fingerprint detection device proposal (Patent Document 6), a space is required between the finger and the imaging optical system, and a finger fixing frame is also required because of the necessity of focusing. And hindered the miniaturization of the device. Further, only the capillary blood vessel exists in the fingerprint portion of the fingertip, and the capillary blood vessel cannot read the pattern by the above method. Since the venous blood vessel that can be read is at the base of the finger below the first joint, it must be read simultaneously with the fingerprint portion of the fingertip above the first joint with a reduction optical system. There is also a problem that it ends up.

“Purpose of invention”
The present invention has been proposed in view of such circumstances, and an object of the present invention is a small size capable of stably inputting biological features such as a fingerprint of a finger using a one-dimensional or quasi-one-dimensional image sensor. An inexpensive biometric feature input device is provided.

  Another object of the present invention is to provide a small and inexpensive biometric feature input device capable of simultaneously inputting a blood vessel image of a finger simultaneously with a fingerprint on the finger surface.

  Still another object of the present invention is to provide an electronic apparatus having a finger travel guide for stably inputting a biometric feature such as a fingerprint of a finger using a one-dimensional or quasi-one-dimensional image sensor. .

  In the first biometric feature input device of the present invention, the one-dimensional or quasi-one-dimensional image sensor and the finger and the effective pixel portion of the image sensor are in contact with each other during relative movement of the finger and the image sensor sliding relative to each other. A finger travel guide that keeps a substantially constant distance without illuminating, and an image by radiation emitted from the skin surface of the finger after being scattered inside the finger by the image sensor during the relative motion. And image processing means for stitching together one-dimensional or quasi-one-dimensional partial images obtained by imaging.

  According to the second biometric feature input device of the present invention, the one-dimensional or quasi-one-dimensional image sensor and the finger and the effective pixel portion of the image sensor are in contact with each other during the relative movement in which the finger and the image sensor slide relative to each other. A finger travel guide that keeps a substantially constant distance without illuminating, an upper light source that irradiates light on the back of the finger with blood vessel image light, and scatters inside the finger from the skin surface of the finger During the relative movement, the image sensor detects a first image by the emitted light and a second image by the emitted light emitted from the skin surface of the finger through the light irradiated from the upper light source. The one-dimensional or quasi-one-dimensional partial images obtained by alternately capturing images are connected to each first image and each second image, and a blood vessel image that is the difference between the reconstructed first image and second image is obtained. With image processing means to extract And wherein the Rukoto.

  According to a third biometric feature input device of the present invention, in the first or second biometric feature input device, the finger travel guide has a gap immediately above the effective pixel portion of the image sensor.

  According to a fourth biometric feature input device of the present invention, in the third biometric feature input device, the height of the gap is 10 μm or more and 200 μm or less, and a width parallel to the relative motion direction is in the sub-scanning direction of the image sensor. The effective pixel length is 2.0 mm or less.

  A fifth biometric feature input device of the present invention is characterized in that, in the fourth biometric feature input device, a light-transmitting solid is inserted in the gap.

  According to a sixth biometric feature input device of the present invention, in the first or second biometric feature input device, at least a portion immediately above the effective pixel portion of the image sensor of the finger travel guide is made of a light transmissive solid. It is characterized by being.

  According to a seventh biometric feature input device of the present invention, in the sixth biometric feature input device, the height of the solid is 10 μm or more and 200 μm or less.

  The eighth biological feature input device of the present invention is the fifth, sixth or seventh biological feature input device, wherein the refractive index of the solid is greater than 1.1.

  According to a ninth biometric feature input device of the present invention, in the fifth, sixth, or seventh biometric feature input device, the refractive index of the solid is larger than 1.1 and smaller than 1.4. .

  According to a tenth biological feature input device of the present invention, in the fifth, sixth, or seventh biological feature input device, a refractive index of the solid is greater than 2.0.

  An eleventh biometric feature input device according to the present invention is the fifth, sixth or seventh biometric feature input device, wherein the solid has a refractive index larger than 2.0 and smaller than 5.0. .

  A twelfth biological feature input device according to the present invention is the first, second, third, fifth, or sixth biological feature input device, wherein light is applied to the abdomen of the finger from the vicinity of a portion to be read by the image sensor. A lower light source for generating scattered light inside the finger by irradiation is provided.

  A thirteenth biological feature input device according to the present invention is a band for extracting an image component of a fingerprint pitch from an output image signal of the image sensor in the first, second, third, fifth, or sixth biological feature input device. A pass filter and an automatic gain control circuit for amplifying the output of the band pass filter are provided.

  A fourteenth biometric feature input device according to the present invention is the first, second, third, fifth or sixth biometric feature input device, wherein the image processing means calculates the distortion of the stitched image at the frequency of the fingerprint portion. A correction means for correcting by analysis is included.

  The first electronic device according to the present invention has a height of 10 μm or more and 200 μm or less immediately above the effective pixel portion of a one-dimensional or quasi-one-dimensional image sensor, and a short side width that is effective in the sub-scanning direction of the image sensor. There is a gap that is not less than the pixel length and not more than 2.0 mm, and the finger and the effective pixel portion of the image sensor do not contact each other during the relative movement in which the finger and the image sensor slide relative to each other. It is characterized by comprising a finger travel guide for keeping it.

  According to a second electronic device of the present invention, in the first electronic device, a light-transmitting solid is inserted in the gap.

"Action"
In the present invention, during the relative movement in which the finger and the one-dimensional or quasi-one-dimensional image sensor slide relative to each other, the finger travel guide prevents contact between the finger and the effective pixel portion of the image sensor. Since the distance between the finger and the effective pixel portion is too long, the image is not blurred and the distance is fluctuated and the image is distorted. The radiated light emitted from the skin surface can be stably imaged by the image sensor during the relative motion, and thus the finger generated by joining the captured one-dimensional or quasi-one-dimensional partial images. The accuracy of the entire image can be increased.

  According to the present invention, the following effects are produced.

  The first effect is that a biometric feature such as a fingerprint of a finger can be stably input by an image sensor. The reason for this is that although it is difficult to obtain a fingerprint image stably with a two-dimensional image sensor being kept in a non-contact state at a certain distance with respect to a finger with curvature, the present invention is one-dimensional or quasi-one-dimensional. And the finger and the effective pixel portion of the image sensor do not come into contact with each other during the relative movement in which the finger and the one-dimensional or quasi-one-dimensional image sensor slide relative to each other. This is because a finger travel guide is provided so as to be maintained.

  The second effect is that biometric features can be input even under adverse conditions such as finger wetting and drying, and skin peeling due to dermatitis. The reason for this is that the finger and the image sensor are kept in contact with each other while maintaining an appropriate distance and a non-contact state so that the image can be directly read by the relative movement between the finger and the image sensor arranged in proximity on the one-dimensional or quasi-one-dimensional image sensor. This is because it is possible to read a finger image reflecting the structure. Further, this effect will be described using an actual image example. FIG. 9 is an image of a method using the total reflection critical angle among methods based on the conventional contact, and the image is missing at a part where the skin is peeled off at a substantially central portion of the image. FIG. 8 is an image example of the same part of the same finger according to the embodiment of the present invention. Contrast is obtained at the same skinned area. FIG. 13 is an example of an image of the same part according to another embodiment of the present invention. In this image as well, the image is obtained without being lost although the brightness is reversed with respect to the peeled part.

  A third effect is to provide a device capable of inputting a small and inexpensive biometric feature. The reason is that the finger and the image sensor are read while maintaining a non-contact state and an appropriate distance at which an image can be directly read by the relative movement between the finger and the image sensor arranged in proximity on the one-dimensional or quasi-one-dimensional image sensor. This is because by adopting a method of joining finger images reflecting the structure of the inside of the finger, parts such as an image sensor can be reduced in size and can be used at low cost.

  The fourth effect is that a biometric feature input device with higher accuracy than the conventional biometric feature input device using only a fingerprint can be provided. The reason is that by moving the finger, not only the fingerprint of the fingertip but also the pattern and blood vessel image between the first joint and the second joint of the finger can be input and more biological features can be read simultaneously.

“First Embodiment”
Next, a first embodiment of the present invention will be described in detail with reference to the drawings.

"Configuration Description"
Referring to FIG. 1, the biometric feature input device according to the first exemplary embodiment of the present invention includes a one-dimensional or quasi-one-dimensional image sensor 5 and a gap 2 immediately above the effective pixel portion 1 of the image sensor 5. From the finger travel guide 3 attached to be positioned, the A / D converter 7 that converts the analog output signal of the image sensor 5 into a digital signal, the control of the imaging timing of the image sensor 5 and the A / D converter 7 And a microprocessor unit 8 that executes image processing on the output digital signal.

  The one-dimensional image sensor 5 is a one-line image sensor, and the quasi-one-dimensional image sensor 5 is a strip-shaped image sensor of about 2 to 20 lines. Considering the specifications, among the biometric features of the finger, the fingerprint ridge spacing is about 0.2 mm to 0.5 mm for adults and 0 for children and women to be able to read fingerprints. Therefore, the pitch of the sensors (light receiving elements) is preferably about 20 to 50 μm. Considering the width and roundness of the finger, if the effective width is about 15 mm in width, for example, if a sensor with 1 line and 512 dots is arranged at 29.6 μm intervals to make a quasi-one-dimensional image sensor at a time, A strip-shaped image having a width of 15.15 mm and a length of 0.35 mm is obtained. The image sensor 5 can be produced by CMOS, CCD, TFT technology, etc. This density and size can be sufficiently produced by the current integrated circuit technology, and an image sensor put to practical use in a video camera or the like is 10 μm. Considering the following, sufficient sensitivity can be obtained.

  The finger travel guide 3 is maintained at a substantially constant distance without contact between the finger 4 and the effective pixel portion 1 of the image sensor 5 during the relative movement during fingerprint collection in which the finger 4 and the image sensor 5 slide with each other. Thus, it is provided between the finger 4 and the image sensor 5. In the case of the present embodiment, the finger travel guide 3 is made of an opaque material and is attached to a housing (not shown) dedicated to the biometric feature input device or attached to a housing of an electronic device such as a mobile phone or a personal computer. Alternatively, it constitutes a part of the casing of such an electronic device. The shape of the gap 2 provided in the finger travel guide 3 of the present embodiment is rectangular when viewed from directly above, and the long side size is such that light is applied to the effective pixel portion 1 of the image sensor 5. There is at least a size larger than the long side (main scanning direction) of the effective pixel unit 1 so as to be sufficiently incident, and the short side size is at least as short as the effective pixel unit 1 so that light is sufficiently incident on the effective pixel unit 1. If the size is larger than the side (sub-scanning direction) and is too large, the skin of the finger 4 will be in direct contact with the effective pixel portion 1 during fingerprint collection, so that the short side is 2.0 mm or less, preferably 1.0 mm or less. It is a size. Further, if the size of the gap 2 in the height (depth) direction is too small, the skin of the finger 4 directly contacts the effective pixel unit 1 during fingerprint collection, and if too large, the skin of the finger 4 and the effective pixel unit 1 Since the image is too far away and the blur of the image becomes severe, the size is 10 μm to 200 μm, preferably 20 μm to 80 μm.

  The A / D conversion unit 7 converts the analog output signal of the image sensor 5 into a digital signal and outputs the digital signal to the microprocessor unit 8, and the microprocessor unit 8 inputs the digital signal of the A / D conversion unit 7 appropriately. Image processing is executed.

"Description of operation"
When reading the fingerprint of the finger 4 using the biometric feature input device of the present embodiment, the first joint of the finger 4 is placed near the gap 2 of the finger travel guide 3 and the gap 2 is traced by the abdomen of the finger 4. In this manner, the finger 4 is pulled in the direction of the arrow 6 in FIG. Since the skin of the finger 4 is elastic, when the finger 4 is pressed in the direction of the arrow 601 as shown in FIG. 2, the height and width are such that the finger 4 comes into contact with the effective pixel portion 1 of the image sensor 5. Even if it is the gap 2, if the gap 2 is traced lightly on the abdomen of the finger 4 as described above, a force 603 opposite to the pulling direction 602 is applied to the skin surface of the finger 4 as shown in FIG. The finger 4 does not come into contact with the pixel unit 1, and the distance between the skin of the finger 4 and the effective pixel unit 1 is always kept constant while the finger 4 is moving.

  During the movement of the finger 4, an image by the radiated light scattered from the finger 4 and radiated from the skin surface of the finger 4 is captured by the image sensor 5. Here, the light scattered in the finger 4 and emitted from the skin surface of the finger 4 forms a shadow according to the internal structure of the finger shown in FIG. The tissue inside the finger of the epidermis 1004 has a dermis 1005, and a papillary gland tissue 1003 exists under a ridge 1002 that is a convex portion of a fingerprint. The dermis 1005 including the papillary gland contains more water and oil than the epidermis 1004 and has a difference in refractive index. Therefore, it is considered that the light emitted from the valley 1001 which is the concave portion of the fingerprint is reduced by the papillary gland protruding from the fingerprint ridge. For this reason, among the sensors (light receiving elements) arranged in the effective pixel portion 1 of the image sensor 5, the sensor close to the ridge 1002 at the timing of imaging is incident as compared to the sensor close to the valley 1001. Light is reduced, and a partial image is obtained in which the valley 1001 is a bright area and the ridge 1002 is a dark area.

  The analog signal relating to the one-dimensional or quasi-one-dimensional partial image obtained at appropriate timings in this way is converted into a digital signal by the A / D conversion unit 7 and input to the microprocessor unit 8. The microprocessor unit 8 reconstructs a skin pattern image of the entire finger 4 by connecting sequentially input partial images. The partial image joining process performed by the microprocessor unit 8 is performed by determining the similarity between the partial images, basically in the same manner as the method described with reference to FIG. An example of the processing is shown in FIG.

  First, a partial image for one frame of the image sensor 5 is read and written in a first memory (not shown) (step S101). Here, the partial image for one frame means an image obtained on all lines in the case of the quasi-one-dimensional image sensor in which the image sensor 5 is composed of several lines, and the image sensor 5 is composed of one line. In the case of a one-dimensional image sensor, it means an image obtained by that one line. Next, the partial image for one frame of the image sensor 5 is read again, and compared with the highest line side of the image stored in the first memory in line units (step S102). If there is a line that is different from the line on the most significant line side of the image stored in the first memory among the lines read this time, the line after the line having the difference is added to the most significant line side of the first memory. . For example, the image sensor 5 is composed of 12 lines, and among the 12 lines read this time, the last 3 lines are the same as the 3 lines on the most significant line side stored in the first memory, and the 4th from the end. If the first line differs from the first line, the ninth line to the ninth line are added to the top line side of the first memory. The processes in steps S102 to S104 are repeated until image data for one finger is acquired (step S105).

  Next, the effect of this embodiment will be described.

  According to the present embodiment, a skin pattern image that directly reflects the internal structure of the finger 4 is obtained by using the one-dimensional or quasi-one-dimensional image sensor 5 and without using extra optical components. Reading can be performed stably without being affected by wetness or dryness, and the apparatus can be simply and miniaturized. The reason is that a small and inexpensive one-dimensional or quasi-one-dimensional image sensor is used as the image sensor 5, and the effective pixel portion of the finger 4 and the image sensor 5 during the relative movement of the finger 4 and the image sensor 5 sliding relative to each other. 1 is provided with a finger travel guide 3 for maintaining a substantially constant distance without contact with 1, and an image by radiated light emitted from the skin surface of the finger 4 by being scattered inside the finger 4. This is because the image sensor 5 directly captures images during relative motion, and the obtained one-dimensional or quasi-one-dimensional partial images are connected by image processing of the microprocessor unit 8 to reconstruct a finger pattern image.

  In addition, when a reduction optical system such as Patent Document 6 is used, even in a skinned portion where good contrast is not obtained due to a phenomenon that is diffused on the surface of the skin and is diffused by a lens or an optical path, this embodiment is applied. According to this, a pattern with good contrast reflecting the structure inside the finger can be obtained. This is considered to be because, in the present embodiment, light is directly incident on the image sensor 5 from the finger at a distance close to the finger 4, so that there are fewer components that diffuse and mix on the surface of the skin.

“Second Embodiment”
Referring to FIG. 6, the biometric feature input device according to the second exemplary embodiment of the present invention is the first embodiment shown in FIG. 1 in that a plurality of light sources 151 are arranged on the finger travel guide 3. This is different from the biometric feature input device according to the first embodiment, and the rest is the same as the first embodiment.

  The light source 151 is arranged in a line along the long side in the vicinity of the gap 2 of the finger travel guide 3, and moves the finger 4 moving on the finger travel guide 3 in the direction indicated by the arrow 6 during fingerprint collection on the back side. Illuminate from (finger travel guide 3 side) to generate scattered light inside the finger. The reason why the light source 151 is arranged on the side where the finger 4 is pulled with the gap 2 as the center is to allow the scattered light to be sufficiently generated inside the fingertip even when the fingertip reaches the vicinity of the gap 2.

  The light scattered inside the finger and emitted from the skin surface can read a skin-pattern pattern only with ambient light, but in addition, the light source 151 is placed close to the one-dimensional or quasi-one-dimensional image sensor 5 in the direction in which the finger is pulled. If it arrange | positions, the light from the light source 151 will be scattered inside a finger | toe, and the light component of a light source direction will inject | emitted strongly. This will be described with reference to FIG.

  Referring to FIG. 7, among scattered light whose intensity is biased in the direction of the light source, the light passing through the vicinity of the ridge 1002 indicated by the arrow 1011 is considered to be darker as the distance passing through the epidermis 1004 becomes longer. On the other hand, light passing through the vicinity of the fingerprint valley 1001 indicated by the arrow 1102 is considered to be brighter as the distance passing through the epidermis 1004 is shortened. Therefore, the contrast increases due to the difference in distance. Although the actual detailed mechanism is unknown, the result of the experiment is shown in FIG. 8 as an image example. For reference, FIG. 9 shows an image obtained by reading the same part of the same finger by a method using the total reflection critical angle among methods based on the conventional contact.

  In FIG. 8, the near side (the lower side in the figure) is the light source and the side where the finger is pulled. The farther fingerprint ridges are darker and the front side of the valleys is brighter. In particular, the light source side of the fingerprint ridge is darker and the contrast is increased. This portion is thought to overlap with the attenuation effect of scattered light inside the skin of the finger by the papillary gland tissue 1003. Further, there is a portion where the pattern is missing around the center of the image in FIG. 9, and this corresponds to the part where the skin is peeled off. In the same part, a pattern appears in FIG. 8, and an image of a skinned portion that has been conventionally lost is also obtained with high contrast.

  In this embodiment, the light source 151 is disposed on the side where the finger 4 is pulled with the gap 2 as the center. However, the light source 151 may be disposed on the opposite side with the gap 2 as the center. The light sources 151 may be arranged in the vicinity of both sides.

“Third embodiment”
Referring to FIG. 10, in the biometric feature input device according to the third exemplary embodiment of the present invention, a protective cover 801 made of a light transmissive solid is inserted in the gap 2 of the finger travel guide 3. 1 is the same as the first embodiment except for the biometric feature input apparatus according to the first embodiment shown in FIG.

  The lower surface of the protective cover 801 is substantially in contact with the effective pixel portion 1 of the image sensor 5, and the upper surface thereof is substantially flush with the upper surface of the finger travel guide 3. Therefore, in order to read the fingerprint of the finger 4, the first joint of the finger 4 is placed near the protective cover 801 embedded in the gap 2 of the finger travel guide 3, and the finger 4 is traced by the abdomen of the finger 4. When pulling, a part of the skin of the finger 4 always touches the protective cover. For this reason, light scattered from within the finger and radiated from the surface of the finger skin is emitted directly from the fingerprint ridge or the like in contact with the protective cover as shown by reference numeral 1111 in FIG. The light enters the protective cover 801, propagates through the protective cover 801, and reaches the effective pixel portion 1 of the image sensor 5. In addition, light emitted from a fingerprint valley or the like not in contact with the protective cover 801 is incident on the protective cover 801 after entering the air layer and propagating through the air layer as indicated by reference numeral 1112, and thereafter the fingerprint ridge portion. The light propagates through the protective cover 801 in the same manner as the light emitted from the image sensor 5 and reaches the effective pixel portion 1 of the image sensor 5.

  On the other hand, in the case of the first embodiment in which the protective cover 801 is not provided, the light scattered from the finger and emitted from the skin surface of the finger is shown in FIG. As indicated by reference numerals 1111 and 1112), the light reaches the effective pixel unit 1 after being incident on the air layer and propagating through the air layer.

  In the case of the first embodiment without the protective cover 801 shown in FIG. 11B, as described above, the ridge portion is detected by the image sensor 5 as a dark portion region and the valley portion is detected as a bright portion region. On the other hand, when the protective cover 801 shown in FIG. 11A is interposed, if the refractive index of the protective cover 801 is close to the same value “1” as air, the protective cover 801 does not exist. Therefore, if the value of the refractive index of the protective cover 801 increases, the ridge portion becomes the bright portion region, and the ridge portion is detected as the dark portion region and the valley portion as the bright portion region. The valley is detected by the image sensor 5 as a dark area. This is because when the refractive index of the protective cover 801 increases, the refractive index difference between the finger 4 and the air and between the air and the protective cover 801 are larger than the refractive index difference between the finger 4 and the protective cover 801. One interface (finger and protective cover) with a small difference in refractive index until the light 1111 radiated from the ridge portion in FIG. 11A reaches the effective pixel portion 1 is increased. The light 1112 radiated from the valley portion passes through two interfaces having a large refractive index difference (interface between the finger and air and the interface between air and the protective cover). The intensity of the radiated light in the valley is stronger than that of the ridge, but the intensity of the light arriving from the ridge is higher than that of the valley. Presumably because it became relatively strong. In fact, in the fingerprint input device disclosed in Patent Document 4 in which scattered radiation from a finger is imaged by a two-dimensional image sensor placed close to the finger through a transparent protective cover made of glass or the like, the valley of the fingerprint is a dark region, a ridge. A fingerprint image in which the line part is a bright part region is obtained.

  For this reason, when the refractive index of the protective cover 801 is a certain value, the contrast between the ridge and the valley becomes zero. In this specification, such a refractive index value is referred to as a singular point, and the protective cover 801 is made of a light-transmitting solid having a refractive index other than a value near the singular point. Hereinafter, the refractive index of the protective cover 801 will be considered.

In Patent Document 5 according to the proposal of the present inventor, the relationship between the refractive index and the contrast of the transparent solid film interposed between the two-dimensional image sensor and the finger is analyzed, and as shown in FIG. A relationship has been derived. In FIG. 12, the vertical axis indicates the power of light incident on the transparent solid film immediately below the fingerprint ridge as P _3L , and the power of light incident on the transparent solid film directly below the fingerprint valley as P _{3D. 3L−} P _3D ) / P _3L calculated by contrast and the horizontal axis represents the refractive index of the transparent solid film. Further, the line connecting the points with + marks assumes that the refractive index of the finger is 1.4, and the line connecting the points with × marks is that with the finger having a refractive index of 1.5. However, in the graph of FIG. 12, only the effect due to the difference in refractive index between the finger skin, air, and the interface of the transparent solid film is obtained by calculation, and is different from the effect due to the finger skin structure.

  Referring to FIG. 12, the contrast is 0% when the refractive index of the transparent solid film is 1.0, which is the same as that of air. This is because the graph of FIG. 12 assumes that the power of light traveling from the inside of the skin toward the ridge and the power of light traveling toward the valley are the same. Originally, the same contrast as in the first embodiment can be obtained when the refractive index is 1.0. In FIG. 12, the contrast value is negative. If the contrast obtained in the first embodiment is C%, the refractive index value at which the contrast is C% in the graph of FIG. In general, since C = about 10, the singular point = 1.1, and the contrast between the valley and the ridge is 0 in the protective cover 801 having a refractive index of 1.1. Therefore, the refractive index of the protective cover 801 needs to be 1.0 or more and less than 1.1 or greater than 1.1. Since there is almost no light-transmitting solid having a refractive index of less than 1.1, the protective cover 801 may be made of a light-transmitting solid having a refractive index substantially higher than 1.1.

  On the other hand, referring to FIG. 12, the contrast is particularly high when the refractive index of the transparent solid film is in the range of 1.4 to 2.0. When the whole peeled part does not contact the transparent solid film, the whole part does not have the same contrast, but a pattern reflecting the structure inside the finger appears as described above. For this reason, if the contrast between the ridges that contact the transparent solid film and the valleys that do not contact it is abnormally high compared to the contrast of the pattern, if the dynamic range of the sensor is not wide, the pattern is detected on the skin. Becomes difficult. Therefore, the refractive index in the range of 1.4 to 2.0 where the contrast is particularly high in FIG.

  Further, as analyzed in Patent Document 5 relating to the proposal of the present inventor, when the refractive index of the transparent solid film increases, the brightness decreases even when the contrast appears, and noise caused by disturbance light and noise generated in the circuit are reduced. The S / N ratio decreases as a component, and the probability that the fingerprint ridge portion and the fingerprint valley portion are not accurately identified increases. For this reason, the upper limit of the refractive index is desirably about 5.0.

  As a result of the above consideration, the refractive index of the protective cover 801 is preferably greater than 1.1 and less than 1.4, or greater than 2.0 and less than 5.0.

  As a solid suitable for the protective cover 801 with a material having a refractive index of less than 1.4, for example, there is glass mainly composed of BeF3 (beryllium fluoride).

  As a solid suitable for the protective cover 801 with a refractive index higher than 2.0, for example, glass containing a large amount of BaO (barium oxide) or PbO (lead oxide), hematite (red iron steel), rutile (goldenite), germanium, There are diamond or silicon. Among these, silicon is easily available as a semiconductor material, is easy to process, and is relatively inexpensive. If a silicon wafer is processed to a thickness of 200 μm or less to form a protective cover, the light has a high transmittance in a low wavelength region, particularly near infrared light having a wavelength of 800 to 1000 nm, and a sufficient sensor light output can be obtained. Silicon is also an environmentally friendly material compared to glass containing harmful substances. In addition, the lower part of the image sensor such as a CMOS or CCD made from a silicon wafer is thinly polished so that the thickness up to the photosensitive layer is 200 μm or less and is turned upside down. If they are brought into contact with each other, an equivalent structure can be obtained without any special cover.

  FIG. 13 shows a fingerprint image of the finger 4 read by the biometric feature input device of this embodiment provided with the protective cover 801. It can be seen that the contrast of the ridge is also obtained in the round skinned portion at the upper left of the image. However, the bright part and the dark part are inverted from other places. When the protective cover 801 is provided as described above, depending on the refractive index of the protective cover 801, the fingerprint ridge portion becomes bright and the valley portion becomes dark depending on the condition of the contact portion, and the brightness and darkness are reversed from the non-contact portion. This problem can be solved by image processing and fingerprint authentication. That is, it is only necessary to extract and connect only the continuity of ridges by edge enhancement. In addition, if an authentication method is used in which authentication is performed based on the positional relationship between feature points such as fingerprint branch points and end points, inversion of light and dark will not affect authentication.

  As described above, according to the present embodiment, in addition to obtaining the same effects as those of the first embodiment, there is no concern that dust accumulates in the gap 2 of the finger travel guide 3 and the image quality deteriorates. effective.

“Fourth embodiment”
Referring to FIG. 14, in the biometric feature input device according to the fourth exemplary embodiment of the present invention, a band-pass filter 1801 and an automatic gain control circuit 1802 are connected between the image sensor 5 and the A / D converter 7. In this respect, the biometric feature input apparatus according to the third embodiment shown in FIG. 10 is different from the biometric feature input apparatus according to the third embodiment, and is otherwise the same as the third embodiment.

  The band pass filter 1801 extracts only the image component of the fingerprint pitch from the image signal output from the image sensor 5. The optimum frequency characteristic of the band-pass filter 1801 is determined from the sensor density and the scanning frequency in consideration of the fingerprint ridge pitch of 0.2 mm to 0.5 mm. The image component extracted by the band pass filter 1801 is amplified by a subsequent automatic gain control circuit 1802 and output to the A / D converter 7.

  According to the present embodiment, since the band pass filter 1801 that extracts only the image component of the fingerprint pitch from the output of the image sensor 5 and the automatic gain control circuit 1802 that amplifies the output are provided, the skinned portion is small. The output can also be increased. When a material with a refractive index in the range of 1.4 to 2.0 is used as the protective cover 801, in the third embodiment, the output of the peeled portion becomes too small and recognition becomes difficult. Such a problem can be improved in this form. If a material such as ordinary glass having a refractive index of 1.4 or more and 2.0 or less can be used for the protective cover 801, it is advantageous in price. Of course, this embodiment is also effective when the protective cover 801 is made of a material having a refractive index other than 1.4 to 2.0.

"Fifth embodiment"
Referring to FIG. 15, the biometric feature input device according to the fifth exemplary embodiment of the present invention is different from the third exemplary embodiment in that the entire finger travel guide 3 is a protective cover 901, and other points are as follows. This is the same as in the third embodiment.

  The protective cover 901 is made of a light transmissive solid having the same refractive index as that of the protective cover 801 in the third embodiment, and the thickness condition is the same as that of the protective cover 801.

  According to the present embodiment, in addition to the same effects as those of the third embodiment, since the entire travel guide 3 is the protective cover 901, there is an effect that the assemblability is excellent.

“Sixth embodiment”
Referring to FIG. 16, the biometric feature input device according to the sixth exemplary embodiment of the present invention includes a light source 161 that illuminates the back part (nail side) of the finger 4 from above, and a blood vessel image simultaneously with the skin pattern of the finger 4. Is different from that of the second embodiment, and other points are the same as those of the second embodiment.

  The light source 161 attached to the upper portion of the finger 4 by a support tool (not shown) is for reading the blood vessel of the finger, and has a near-infrared absorption stronger than other biological tissues. Irradiate the light. In particular, an LED developed for an infrared remote controller having a wavelength near 820 to 950 nm has a large output and is suitable for the light source 161. Only an image of the skin surface is obtained from the image derived from the light source 151 disposed at the lower part of the finger 4, but the image by the light source irradiated from the upper light source 161 passes through a thick blood vessel in which blood containing hemoglobin is concentrated. It includes images of blood vessels that are darker than other tissues. Therefore, the blood vessel image can be read simultaneously with the skin pattern of the finger by obtaining an image by the light source 161 and an image by the light source 151 and calculating a difference between the images as follows.

  When the blood vessel image is read simultaneously with the finger skin pattern using the biometric feature input device according to the present embodiment, the second joint of the finger 4 is placed near the gap 2 of the finger travel guide 3 and the abdomen of the finger 4 The finger 4 is pulled so as to trace the gap 2. An image is picked up by the image sensor 5 while the finger 4 is moving. First, when the first frame image of the image sensor 5 is obtained, the light source 151 arranged at the lower part of the finger is turned on to obtain an image. Next, the light source 151 at the lower part of the finger is turned off and the light source 161 at the upper part of the finger is turned on to obtain the next frame image of the image sensor 5. By repeating this and alternately connecting the images of the respective light sources, an image 1701 by the lower light source 151 and an image 1702 by the upper light source 161 as shown in FIG. 17 are obtained. In both cases, a fingerprint 1704 and a pattern 1707 between the first joint 1705 and the second joint 1706 are present, but the image from the light source 161 from the top further includes a blood vessel image 1708. By obtaining the difference between two images 1701 and 1702 that are alternately switched, an image 1703 having only a blood vessel image 1709 can be obtained. The processing for obtaining the difference is performed by the microprocessor unit 8. FIG. 18 shows an example of the processing.

  First, a partial image for one frame of the image sensor 5 is read in a state where only the light source 151 is turned on, and is written in a first memory (not shown) (steps S201 and S202). Next, a partial image for one frame of the image sensor 5 is read in a state where only the light source 161 is turned on, and is written in a second memory (not shown) (steps S203 and S204). Next, the partial image for one frame of the image sensor 5 is read again with only the light source 151 turned on, and compared with the line on the most significant line side of the image stored in the first memory in line units (step S205, S206) If there is a line that is different from the line on the most significant line side of the image stored in the first memory among the lines read this time, the line after the line having the difference is regarded as the most significant line side of the first memory. (Steps S207 and S208). Next, a partial image for one frame of the image sensor 5 is read with only the light source 161 turned on, and compared with the line on the most significant line side of the image stored in the second memory in units of lines (steps S209 and S210). ) If there is a line that is different from the line on the most significant line side of the image stored in the second memory among the lines read this time, the line after the difference line is placed on the most significant line side of the second memory. It adds (steps S211 and S212). The processes in steps S205 to S212 are repeated until image data for one finger is acquired (step S213). Accordingly, the image 1701 in FIG. 17 is stored in the first memory, and the image 1702 in FIG. 17 is stored in the second memory. Finally, the image stored in the first memory is subtracted from the image stored in the second memory to generate the image 1703 of FIG. 17 (step S214).

  As described above, according to the present embodiment, by sliding the finger 4 on the finger travel guide 3 from the second joint 1706 to the fingertip, the finger joint 1704 is moved between the first joint and the second joint at the same time as the fingerprint 1704. The skin pattern 1707 and the blood vessel image 1709 can be read. Unfortunately, because the blood vessel image becomes thin or changes when the finger is pressed, the accuracy of the personal authentication information is insufficient, but it is used as interpolated data for personal authentication by skin patterns such as fingerprints, or for fake fingers. Since it can be used for determination, there is an effect that personal authentication can be performed with higher accuracy than with a fingertip fingerprint alone with all images.

“Seventh embodiment”
The biometric feature input device according to the seventh exemplary embodiment of the present invention is the first exemplary embodiment in that a process for correcting image distortion is executed after the partial image joining process in the microprocessor unit 8. The other points are the same as those of the first embodiment. Therefore, the configuration of the present embodiment is the same as FIG.

  In the present embodiment, the microprocessor unit 8 performs a process for joining the partial images and a process for correcting the distortion of the images in that order. The process of joining the partial images is the same as in the first embodiment. In this case, the image in the lateral direction of the finger is uniquely distorted by the sensor pitch of the image sensor 5, but the vertical direction is distorted by the moving speed of the finger 4 even when connected while looking at the correlation. . Among the authentication methods such as fingerprints, the method of checking the correlation between the branch points and the end points of the ridges is relatively resistant to distortion, but it is still desirable to be corrected in order to improve the authentication accuracy. Therefore, in the present embodiment, the fingerprint has ridge components in the horizontal and vertical directions, and the interval between the ridges is almost constant for each individual. Predict distortion and correct as follows.

In FIG. 19, the horizontal ridge frequency component 1202 of the original finger print image 1201 is f1, the vertical ridge frequency component 1203 is f2, and the ordinate of pixels on the ridge before correction is Y. Then, the corrected pixel ordinate Y ′ on the ridge is given by the following equation.
Y ′ = Y × (f2 / f1) (1)

  FIG. 19 shows a case where the image is stretched vertically by slowly pulling a finger. However, even when the finger is pulled too quickly and shortened, a partial defect is present, but an approximate image can be obtained although there is a limit depending on the degree of defect. FIG. 19 shows a spiral pattern in the form of a fingerprint, and FIG. 20 shows a hoof pattern. Statistically, these two patterns are often included in humans. In rare cases, there is an arcuate pattern as shown in FIG. 21. In the case of this arcuate pattern, the frequency components of the ridges in the horizontal direction are greatly different, and thus the method described above cannot be applied. However, since there are few arcuate patterns (it is said to be 1% or less in Japanese), most pattern images can be corrected by the above correction method.

  FIG. 22 shows a flow of image processing performed by the microprocessor unit 8. Steps S301 to S305 indicate processing for joining the partial images, and are the same as steps S101 to S105 in FIG. Steps S306 to S308 show processing procedures for correcting image distortion. In the process of correcting image distortion, first, ridges are extracted by performing image processing such as edge enhancement and skeleton processing on the image stored in the first memory (step S306). Next, the number of horizontal and vertical ridges of the image is obtained and divided by the number of horizontal and vertical pixels, respectively, thereby obtaining the frequency component f1 of the horizontal ridge and the frequency component f2 of the vertical ridge (step) S307). Next, the shape of the fingerprint pattern is judged from the shape of the ridge, and if it is the spiral pattern and the hoof pattern shown in FIGS. 19 and 20, the ordinate of the pixel on the ridge is calculated using the above-described equation (1). Is corrected and the image is expanded and contracted in the vertical direction (step S308).

  Thus, according to the present embodiment, the same effect as that of the first embodiment can be obtained, and a pattern image with less distortion can be obtained. The reason is that the distortion in the vertical direction of the image is predicted and corrected based on the difference in the frequency components of the ridges in the horizontal and vertical directions in the image reconstructed by connecting the partial images.

  While the present invention has been described with reference to some embodiments, the present invention is not limited to the above embodiments, and various other additions and modifications can be made. For example, an embodiment in which the light source 151 in the second embodiment is provided in the third to seventh embodiments, a fourth embodiment in the first, third, and fifth to seventh embodiments. An embodiment in which the above embodiments are appropriately combined, such as an embodiment in which the band pass filter 1801 and the automatic gain control circuit 1802 are provided, is also conceivable.

  As described above, the biometric feature input device according to the present invention is useful as a small and inexpensive reading device that stably reads a pattern such as a fingerprint of a finger and a blood vessel image, and particularly due to finger wetting and drying, dermatitis, and the like. It is suitable as a device that can input biometric features even under adverse conditions such as skin peeling.

It is the top view and cross-sectional view of the biological feature input device concerning the 1st Embodiment of this invention. It is a figure explaining the state which pressed the finger | toe into the gap | interval of the finger | toe travel guide of the biometric feature input device concerning the 1st Embodiment of this invention. It is a figure explaining the state which moved the finger along the finger travel guide of the biological feature input device concerning a 1st embodiment of the present invention. It is a figure explaining the internal structure of the skin of a finger. It is a flowchart which shows the process example by the microprocessor part of the biometric feature input device concerning the 1st Embodiment of this invention. It is the upper side figure and cross-sectional view of the biometric feature input device concerning the 2nd Embodiment of this invention. It is a figure explaining the effect | action by the light source of the biometric feature input device concerning the 2nd Embodiment of this invention. It is a figure which shows the example of a fingerprint image read with the biometric feature input device concerning the 2nd Embodiment of this invention. It is a figure which shows the example of a fingerprint image input with the conventional fingerprint input device using the total reflection critical angle. It is a cross-sectional view of the biological feature input device concerning the 3rd Embodiment of this invention. It is explanatory drawing of an effect | action of the biometric feature input device concerning the 3rd Embodiment of this invention. It is a graph which shows the relationship between the refractive index and contrast of a transparent solid film interposed between a two-dimensional image sensor and a finger. It is a figure which shows the example of a fingerprint image read with the biometric feature input device concerning the 3rd Embodiment of this invention. It is a cross-sectional view of the biological feature input device concerning the 4th Embodiment of this invention. It is a cross-sectional view of the biometric feature input apparatus according to the fifth embodiment of the present invention. It is a cross-sectional view of the biometric feature input apparatus according to the sixth embodiment of the present invention. It is a figure explaining the principle which reads a blood vessel image simultaneously with a fingerprint image by the biometric feature input device concerning the 6th Embodiment of this invention. It is a flowchart which shows the process example by the microprocessor part of the biometric feature input device concerning the 6th Embodiment of this invention. It is a figure explaining the image correction method with respect to a spiral fingerprint in the biometric feature input device concerning the 6th Embodiment of this invention. It is a figure explaining the image correction method with respect to a hoof-like fingerprint in the biometric feature input device concerning the 6th Embodiment of this invention. It is a figure explaining the shape of an arcuate fingerprint. It is a flowchart which shows the process example by the microprocessor part of the biometric feature input device concerning the 7th Embodiment of this invention. It is a figure explaining the principle of the optical prism system on the assumption of the conventional contact. It is a figure explaining the conventional image reconstruction system. It is a figure explaining the problem at the time of moving a finger late by the conventional image reconstruction. It is a figure explaining the problem at the time of moving a finger early by the conventional image reconstruction.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Light in the direction totally reflected by the prism surface through the air 102 ... Light in the direction passing through the prism surface through the finger skin 104 ... Finger skin 105 ... Prism 106 ... Imaging lens 107 ... Two-dimensional image sensor 108 ... Prism Fingerprinting surface 109 of the prism 1. Fingerprint image exit surface 1... Effective pixel portion 2 of the one-dimensional or quasi-one-dimensional image sensor 2. Finger travel guide gap 3. Finger travel guide 4. Dimensional image sensor 6 ... Finger pulling direction 7 ... A / D converter 8 ... Microprocessor unit 301 ... Original finger image 302 ... Correctly reconstructed finger image 303 ... Long distorted reconstructed finger image 304 Image of finger I1 that is short and distorted and reconstructed Image In at the start of partial image reading In image 601 at the end of partial image reading Just touched without dragging Force direction 602 ... protective cover 1001 ... skin pattern valleys serving as both a direction 801 ... and the protective cover 901 ... finger travel guide which has been subjected to the finger travel guide force when pulling the finger (recess)
1002 ... Skin pattern ridge (convex)
1003 ... Papillary gland tissue 1004 of skin internal tissue ... Skin epidermis part 1005 ... Skin dermis part 1101 ... Bias scattered light from light source passing through fingerprint ridge 1102 ... Bias from light source passing through fingerprint valley Scattered light 1201 ... Distorted and reconstructed spiral fingerprint image 1202 ... Example of location 1203 where the horizontal frequency component of the fingerprint is measured 1203 ... Example of location where the vertical frequency component of the fingerprint is measured 1205 ... Correctly corrected spiral fingerprint Image 1301 ... Distorted and reconstructed hoof-like fingerprint image 1302 ... Example of location 1303 for measuring frequency component of horizontal side of fingerprint ... Example of location for measuring vertical frequency component of fingerprint 1305 ... Correctly corrected hoof-like fingerprint Image 1401... Example of arcuate fingerprint 151... Contrast increasing light source 161 that irradiates the abdomen of the finger... Blood vessel image light source 1701 that irradiates the back of the finger. 1 image 1702 ... image 1703 including blood vessels by the upper light source 161 ... image 1704 extracted blood vessel after image processing ... fingerprint part 1705 ... first joint 1706 ... second joint 1707 ... between the first joint and the second joint Skin pattern 1708 ... Blood vessel image 1709 overlapping with the skin pattern ... Blood vessel image 1801 extracted by image processing ... Bandpass filter 1802 ... Automatic gain control circuit

Claims (16)

  During the relative movement in which the one-dimensional or quasi-one-dimensional image sensor and the finger and the image sensor slide relative to each other, the finger and the effective pixel portion of the image sensor are not in contact with each other so that a substantially constant distance is maintained. A one-dimensional or quasi-one-dimensional image obtained by capturing an image of the finger running guide to be radiated and the emitted light scattered from the finger skin surface and emitted from the finger skin surface during the relative movement by the image sensor. A biometric feature input device comprising image processing means for joining partial images together.   During the relative movement in which the one-dimensional or quasi-one-dimensional image sensor and the finger and the image sensor slide relative to each other, the finger and the effective pixel portion of the image sensor are not in contact with each other so that a substantially constant distance is maintained. A finger travel guide, an upper light source that irradiates the back of the finger with light for blood vessel images, a first image by the radiated light scattered inside the finger and emitted from the skin surface of the finger, and the upper part One-dimensional or quasi-obtained image obtained by alternately imaging the second image by the radiated light emitted from the skin surface of the finger through the light irradiated from the light source by the image sensor during the relative motion. Image processing means for connecting one-dimensional partial images for each first image and each second image, and extracting a blood vessel image that is a difference between the reconstructed first image and the second image, Biometric feature input device .   The biological feature input device according to claim 1, wherein the finger travel guide has a gap immediately above the effective pixel portion of the image sensor.   The height of the gap is 10 μm or more and 200 μm or less, and the width parallel to the relative movement direction is an effective pixel length in the sub-scanning direction of the image sensor and 2.0 mm or less. Biometric feature input device.   The biological feature input device according to claim 4, wherein a light-transmitting solid is inserted into the gap.   3. The biometric feature input device according to claim 1, wherein at least a portion immediately above the effective pixel portion of the image sensor of the finger travel guide is made of a light-transmitting solid.   The biological feature input device according to claim 6, wherein a height of the solid is 10 μm or more and 200 μm or less.   The biological feature input device according to claim 5, wherein the refractive index of the solid is greater than 1.1.   The biological feature input device according to claim 5, 6 or 7, wherein the refractive index of the solid is larger than 1.1 and smaller than 1.4.   The biological feature input device according to claim 5, 6 or 7, wherein the refractive index of the solid is larger than 2.0.   The biological feature input device according to claim 5, 6 or 7, wherein the refractive index of the solid is larger than 2.0 and smaller than 5.0.   The lower light source which generates scattered light inside a finger | toe by irradiating light to the abdominal part of the finger | toe from the vicinity of the site | part to be read by the image sensor is provided. The biometric feature input device described.   A bandpass filter that extracts an image component of a fingerprint pitch from an output image signal of the image sensor, and an automatic gain control circuit that amplifies the output of the bandpass filter. The biometric feature input device according to 5 or 6.   7. The biometric feature input device according to claim 1, wherein said image processing means includes correction means for correcting distortion of the joined images by frequency analysis of a fingerprint portion.   Immediately above the effective pixel portion of the one-dimensional or quasi-one-dimensional image sensor, the height is not less than 10 μm and not more than 200 μm, and the short side width is not less than the effective pixel length in the sub-scanning direction of the image sensor and not more than 2.0 mm. Finger travel for maintaining a substantially constant distance without contact between the finger and the effective pixel portion of the image sensor during relative movement of the finger and the image sensor sliding relative to each other with a gap Electronic equipment with a guide.   16. The electronic apparatus according to claim 15, wherein a light-transmitting solid is inserted in the gap.