WO2008105548A1 - 情報抽出方法、情報抽出装置、プログラム、登録装置及び照合装置 - Google Patents
情報抽出方法、情報抽出装置、プログラム、登録装置及び照合装置 Download PDFInfo
- Publication number
- WO2008105548A1 WO2008105548A1 PCT/JP2008/053714 JP2008053714W WO2008105548A1 WO 2008105548 A1 WO2008105548 A1 WO 2008105548A1 JP 2008053714 W JP2008053714 W JP 2008053714W WO 2008105548 A1 WO2008105548 A1 WO 2008105548A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- living body
- stereoscopic image
- body part
- finger
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
- G06T7/55—Depth or shape recovery from multiple images
- G06T7/564—Depth or shape recovery from multiple images from contours
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/107—Static hand or arm
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/14—Vascular patterns
- G06V40/145—Sensors therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/14—Vascular patterns
Definitions
- the present invention relates to an information extraction method, an information extraction device, a program, a registration device, and a collation device, and is suitable for application to biometric authentication, for example.
- Background art
- Biometric authentication is a technique for identifying whether or not the person is the person using the identification target of the living body.
- One of the living body identification targets is the finger blood vessel.
- an authentication device has been proposed in which a three-dimensional image is generated by combining images of different surfaces of a fingertip and used as an identification target (see, for example, Patent Document 1).
- Patent Literature 1 Patent 2 0 0 2— 1 7 5 5 2 9 Publication.
- 3D images that have a much larger amount of information than 2D images are targeted for identification, which has the advantage of improving the accuracy of identifying the identity of the specific person (person), that is, the accuracy of authentication. is there.
- the present invention is an information extraction method, wherein a common part of a silhouette in a living body part reflected in the image from a plurality of first images with a viewpoint of the surrounding of the living body part as a target space
- the common part of Siletsu ⁇ in the inner surface image of the living body part shown in the image from the multiple second images with the viewpoint around the living body part in the target space The value representing the shape of the cross section of the second stereoscopic image having a predetermined positional relationship with respect to the reference position of the first stereoscopic image and the first step generated as the two-dimensional solid image is extracted as identification information
- a second step is provided.
- the present invention is also an information extraction device having a work memory and an image processing unit that executes image processing in the work memory, the image processing unit comprising: From this image, a common part of the silhouette of the living body part reflected in the image is generated as a first stereoscopic image in the target space, and 'the plurality of second images from the viewpoint of the surrounding of the living body part are A common part of the silhouette in the internal image of the living body image shown in Fig. 2 is generated as a second stereoscopic image in the target space, and the second stereoscopic image having a predetermined positional relationship with the reference position of the first stereoscopic image is generated.
- the value representing the shape in the cross section was extracted as identification information.
- the present invention is a program that targets a common part of a silhouette in a living body part that appears in the image from a plurality of first images with a viewpoint around the living body part, to a control unit that controls a work memory.
- a first stereoscopic image is generated in space, and a common part of Siletsu ⁇ ⁇ ⁇ ⁇ in the internal image of the living body part that appears in the image is displayed on the target space.
- a value representing the shape in the cross section of the second stereoscopic image having a predetermined positional relationship with respect to the reference position of the first stereoscopic image, Extracting as identification information was executed.
- the present invention is a registration device, and generates, from a plurality of first images having a viewpoint around a living body part, a common part of a silhouette in a living body part reflected in the image as a first stereoscopic image in a target space. And generating a common part of the silhouette in the inner surface image of the living body part shown in the image as a second stereoscopic image in the target space from a plurality of second images with the view around the living body part Means for extracting a value representing the shape of the cross section of the second solid image having a predetermined positional relationship with respect to the reference position of the first stereoscopic image, and the value as identification information in the storage medium. And a registration means for registration.
- the present invention is a collation device, which generates a common part of a silhouette in a living body part reflected in the image as a first three-dimensional image in a target space from a plurality of first images having a viewpoint around the living body part.
- a common part of the silhouette in the inner surface image of the living body part shown in the image is generated as a second three-dimensional image in the target space from a plurality of second images with the view around the living body part.
- An extraction means for extracting a value representing the shape of the cross section of the second solid image having a predetermined positional relationship with respect to the reference position of the first stereoscopic image, and a value and identification information registered in the storage medium
- a collation means for collating with the value to be checked is provided.
- the identification information is extracted as representing the shape of the portion having a certain relationship with the first stereoscopic image even if the partial shape itself in the second stereoscopic image is represented.
- Both of the three-dimensional image and the second three-dimensional image can be discretely expressed by some shape values in the second three-dimensional image.
- the first three-dimensional image and the second three-dimensional image are simply displayed.
- an information extraction method, an information extraction device, a program, a registration device, and a verification device that can improve authentication accuracy while suppressing the amount of information to be identified can be realized.
- FIG. 1 is a block diagram showing a configuration of an authentication apparatus according to this embodiment.
- FIG. 2 is a schematic diagram showing the state transition of a rotating finger.
- FIG. 3 is a schematic diagram showing the relationship between the imaging target and the image.
- FIG. 4 is a block diagram showing a functional configuration of the control unit.
- FIG. 5 is a schematic diagram for explaining the detection of a finger joint.
- FIG. 6 is a schematic diagram for explaining the calculation of the rotation correction amount.
- FIG. 7 is a schematic diagram showing blood vessel images before and after embossing.
- FIG. 8 is a schematic diagram for explaining the calculation of the motion amount.
- FIG. 9 is a schematic diagram showing the luminance state of the blood vessel image after the relief processing.
- FIG. 10 is a schematic diagram showing luminance state transition in the blood vessel extraction processing.
- FIG. 11 is a schematic diagram for explaining the uniformization of the luminance state.
- Fig. 12 is a schematic diagram showing the poxel space.
- FIG. 13 is a schematic diagram for explaining the detection of the finger silhouette region.
- FIG. 14 is a schematic diagram for explaining the arrangement relationship of each image arranged around the poxel space.
- FIG. 15 is a schematic diagram showing a finger three-dimensional image.
- FIG. 16 is a schematic diagram for explaining the detection (1) of the blood vessel region.
- Figure 17 shows the relationship between the object and the object image shown in the image.
- FIG. 18 is a schematic diagram for explaining the detection (2) of the blood vessel silhouette region.
- FIG. 19 is a schematic diagram showing the extraction state of the blood vessel region.
- FIG. 20 is a schematic diagram showing a blood vessel stereoscopic image.
- FIG. 21 is a schematic diagram showing a finger stereoscopic image and a blood vessel stereoscopic image generated in the poxel space.
- FIG. 22 is a schematic diagram for explaining the determination of the cross section based on the joint.
- FIG. 23 is a schematic diagram for explaining the extraction of the cross-sectional shape value.
- FIG. 1 shows the overall configuration of authentication apparatus 1 according to the present embodiment.
- This authentication device 1 connects an operation unit 1 1, an imaging unit 1 2, a memory 1 3, an interface 1 4, and a notification unit 1 5 to a control unit 10 via buses 16, respectively. It is constituted by.
- the control unit 10 includes a central processing unit (CPU) that controls the entire authentication device 1, a read only memory (R 0 M) that stores various programs and setting information, and the work memory of the CPU. It is configured as a computer that includes RAM (Random Access Memory).
- CPU central processing unit
- R 0 M read only memory
- RAM Random Access Memory
- execution command C OM 1 of the user to be registered (hereinafter referred to as a registrant) or a registrant himself or herself is registered.
- Execution command COM 2 of a mode for determining whether or not there is present (hereinafter referred to as an authentication mode) is input from the operation unit 11 in response to a user operation.
- the control unit 10 determines a mode to be executed based on the execution instructions C 0M 1 and C OM 2, and based on the program corresponding to the determination result, the imaging unit 1 2, the memory 13, and the The face 14 and the notification unit 15 are appropriately controlled to execute the blood vessel registration mode or the authentication mode.
- the imaging unit 12 Based on the exposure value (EV (Exposure Value) value) specified by the control unit 10, the imaging unit 12 is based on the lens position in the optical system, the aperture value of the aperture, and the shirt speed of the imaging device (exposure time). )
- the imaging unit 1 2 is configured to rotate the subject image displayed on the imaging surface of the imaging device for a predetermined period. Images are taken every period, and data relating to the images generated as the imaging results (hereinafter referred to as image data) are sequentially output to the control unit 10.
- the imaging unit 12 drives the near-infrared light source during the period specified by the control unit 10 so that the imaging target space (hereinafter referred to as imaging space) becomes a blood vessel. Irradiate specifically absorbed near infrared light.
- the memory 13 is, for example, a flash memory, and stores or reads the data specified by the control unit 10.
- the notification unit 15 includes a display unit 15 a and an audio output unit 15 b, and the display unit 15 5 a displays characters and figures based on the display data given from the control unit 10 on the display screen. To do.
- the audio output unit 15 b is configured to output audio based on the audio data given from the control unit 10 from the speaker.
- the control unit 10 shifts the operation mode to the blood vessel registration mode, places a finger on the imaging space, and places the finger on the finger peripheral surface ( The notification unit 15 is notified that it must be rotated along the finger pad, finger side, and finger back surface.
- the control unit 10 causes the imaging unit 12 to perform an imaging operation, and turns on the near-infrared light source in the imaging unit 12 for a predetermined lighting period at predetermined intervals, whereby a near-infrared light source Blinks.
- a finger image when a finger arranged in the imaging space is rotated along the finger peripheral surface, for example, as shown in FIG. (Hereinafter referred to as a finger image) and an image of a blood vessel inside the finger (hereinafter referred to as a blood vessel image) are alternately obtained corresponding to the blinking operation of the near-infrared light source.
- the finger image is picked up using visible light as imaging light
- the blood vessel image is picked up using near infrared light as imaging light.
- control unit 10 based on the image data input from the imaging unit 12 in the imaging order, the three-dimensional image of the finger (hereinafter referred to as the finger three-dimensional image) and the three-dimensional image of the blood vessel.
- a blood vessel stereoscopic image (Hereinafter referred to as a blood vessel stereoscopic image) and a value representing a cross-sectional shape in the stereoscopic image (hereinafter referred to as a cross-sectional shape value) is extracted. Then, the control unit 10 registers the cross-sectional shape value by storing it in the memory 13 as the identification target data (hereinafter referred to as the identification data).
- control unit 10 can execute the blood vessel registration mode.
- the control unit 10 changes the operation mode to the authentication mode, and rotates the finger along the circumferential surface of the finger in the imaging space, as in the blood vessel registration mode.
- the notification unit 15 notifies that it has to be performed, and the imaging unit 1 2 is caused to perform an imaging operation, and the near infrared light source is caused to blink.
- control unit 10 extracts the cross-sectional shape values in the finger stereoscopic image and the blood vessel stereoscopic image in the same manner as in the blood vessel registration mode, based on the image data input from the imaging unit 12 in the imaging order. Then, the control unit 10 collates the extracted cross-sectional shape value with the cross-sectional shape value stored in the memory 13 as the identification data, and from this collation result, whether or not the registrant can approve it. It is made to judge whether or not. If it is determined that it cannot be approved as a registrant,
- the control unit 10 notifies the fact visually and audibly via the display unit 15 a and the audio output unit 1.5 b.
- the control unit 10 is connected to the data base 14 for the data indicating that the user is approved as a registrant.
- the data indicating that the user has been approved as a trigger is used as a trigger, for example, a predetermined process to be executed when authentication is successful, such as locking the door for a certain period of time or canceling the restricted operation mode. Done.
- control unit 10 can execute the authentication mode.
- this processing is performed by a finger joint detection unit 21, an image rotation unit 22, an image extraction unit 23, a blood vessel extraction unit 24, a motion amount calculation unit 25, and a 3D image generation unit. 26 and shape extraction unit 27.
- the finger joint detection unit 21, the image rotation unit 22, the image extraction unit 23, the blood vessel extraction unit 24, the motion amount calculation unit 25, the 3D image generation unit 26, and the shape extraction unit 27 will be described in detail.
- the finger joint detection unit 21 acquires the finger image data DF a,. Finger joint data is detected from a finger image based on DF ai. In addition, when detecting a joint, the finger joint detection unit 21 provides a position data DP i indicating the position of the joint to the image rotation unit 22, the image extraction unit 23, and the shape extraction unit 27.
- the image rotation unit 22 is given DF bi (finger image data) DF bi regarding the finger image from which the finger region is extracted, which is obtained in the joint detection process.
- the finger joint detection unit 21 acquires the finger image data DF ai, as shown in FIG. 5, for example, based on the contrast of the finger image (FIG. 5 (A)), the finger region is detected from the finger image. Is extracted (Fig. 5 (B)).
- the finger joint detection unit 21 extracts points constituting the finger contour (hereinafter referred to as finger contour points) from the finger region using the contour extraction filter (FIG. 5 (C )) From the finger contour points, the finger contour points corresponding to the horizontal direction are extracted by stretching using Hough transform or the like (FIG. 5 (D)).
- the finger joint detector 21 detects a line segment passing through the approximate center of each stretched finger outline as a joint J N L (FIG. 5 (E)). (4-2) Image rotation correction.
- the image rotation unit 22 When the image rotation unit 22 acquires the finger image DF bi DFbi, the image rotation unit 22 recognizes the joint position from the position data DP i corresponding to the finger image DF bi DF bi, and uses the position of the joint as a reference. Rotate and correct the finger image. Then, the image rotation unit 22 gives data (finger image data) DF C i related to the rotation-corrected finger image to the image cutout unit 23.
- the image rotation unit 22 acquires the blood vessel image data DB aj, the position image corresponding to the finger image data DF b A derived immediately before the tube image data DB aj DP i Then, the position of the joint is recognized, and the blood vessel image is rotationally corrected based on the position of the joint. Then, the image rotation unit 22 provides the image cutout unit 23 with the data DBbj regarding the rotation-corrected blood vessel image (blood vessel image data).
- An example of the rotation method in the image rotation unit 22 will be described. For example, as shown in FIG. 6, the image rotation unit 22 obtains the angle 0x formed from the joint JNL with respect to the line LN in the column direction of the image as the rotation correction amount of the finger image.
- the finger image is taken as an example in FIG. 6, the same applies to the blood vessel image.
- the image rotation unit 22 is configured to rotationally correct the image in accordance with the rotation correction amount (S x).
- the finger image and blood vessel image at each viewpoint are rotationally corrected so that the longitudinal direction of the finger in the image is the row direction of the image.
- the angle between the row direction of the image and the extension direction of the joint J N L is 0 [:.
- the angle between the row or column direction of the image and the extension direction of the joint may be a predetermined angle.
- the cut-out section 2 3 recognizes the position of the joint from the position image corresponding to the finger image DF bi, and the position of this joint. As a reference, an area of a predetermined size is cut out from the finger image. Then, the image cutout unit 23 gives data relating to the image of the cutout region (hereinafter referred to as finger image partial data) DFdi to the motion amount calculation unit 25 and the 3D image generation unit 26.
- the image cutout unit 23 acquires the blood vessel image data DB bj, the blood vessel image data.
- the joint image from the position DP DP i corresponding to the finger image data DF c ⁇ acquired immediately before DB bj A region of a predetermined size is cut out from the blood vessel image using the position of the joint as a reference. Then, the image cutout unit 23 provides the blood vessel extraction unit 24 with data relating to the image of the cutout region (hereinafter referred to as blood vessel image partial data) DBC j.
- the blood vessel extraction unit 24 acquires the blood vessel image partial data DB c
- the blood vessel imaged in the blood vessel image is converted to a differential filter such as a Gaussian fill or L 0 g fill.
- the image data is drawn using the image data, and the data relating to the image in which the blood vessel is displayed (the blood vessel image portion data) DBdj is given to the motion amount calculation unit 25.
- Fig. 7 shows an example of images before and after the relief. Blood vessel image before relief
- Fig. 7 (A) the boundary between the blood vessel and the other part is unclear, but in the blood vessel image after the relief (Fig. 7 (B)), the boundary is clear.
- the blood vessel is emphasized by this processing, and as a result, the blood vessel and other parts can be clearly distinguished.
- the blood vessel extraction unit 24 converts the image in which the blood vessel is embossed into a binary image (hereinafter referred to as a binary blood vessel image) based on the set luminance value when the blood vessel is embossed. Then, the data regarding the binary blood vessel image (blood vessel image partial data) DB ej is given to the 3D image generation unit 26.
- the motion amount calculating unit 25 selects the finger image based on the selected finger image partial data DF di, the selection The amount of movement of the finger in the finger image based on the finger image partial data DF di input before the finger image partial data DF di is calculated.
- the motion amount calculation unit 25 calculates the motion amount
- the motion amount calculation unit 25 (hereinafter referred to as motion amount data) D FMH D FM 3 — 5 , DFM 5-7 ,... ... or D FM ( nn is given to the 3D image generator 26.
- the motion amount calculation unit 25 selects the blood vessel image part DB dj input from the image cutout unit 23 as a processing target, the blood vessel image based on the selected blood vessel image partial data DB dj is selected. And the amount of movement of the blood vessel reflected in the blood vessel image based on the blood vessel image partial data DB d input before the selected blood vessel image partial data DB dj.
- the motion amount calculation unit 25 calculates the motion amount data DBM 2 — 4 , DBM 4 — 6 , DBM 6 — 8 ,..., Or DBM ( n 3 ) D as a three-dimensional image. This is given to the generator 26.
- the motion amount calculation unit 25 calculates the motion amount by optical flow.
- a finger image or blood vessel image selected as a processing target is referred to as a current image
- a finger image or blood vessel image input before the finger image or blood vessel image is referred to as a previous image. That is, for example, as shown in FIG. 8 (A), the motion amount calculation unit 25 determines a point of interest (hereinafter, referred to as a point of interest) AP in the current image IM 1, and centered on the point of interest AP.
- Yes (mxn) Block of pixels (Hereafter, this is referred to as the attention block) Recognizes the luminance value in AB L.
- the motion amount calculation unit 25 searches the previous image IM 2 for a block that minimizes the difference from the luminance value in the block of interest ABL.
- the center of RB L is the point corresponding to the target point AP (hereinafter referred to as the corresponding point) XP, and the position vector V to the corresponding point XP with reference to the position AP ′ corresponding to the target point AP ( V, x , V y ).
- the motion amount calculation unit 25 searches the previous image IM 2 for blocks corresponding to the plurality of blocks of interest in the current image IM 1, and the center (XP) of the block and the center of the block of interest And the same position (AP ') for each position vector (horizontal vector component V x average, vertical vector component V y average)' It is made to come out.
- This amount of movement is not only in the direction of the horizontal movement (rotation direction) with respect to the surface on which the finger is placed, but also in the direction of vertical movement with respect to the surface due to finger pressure and rotation axis fluctuation This value represents the movement in the direction (perpendicular to the rotation direction).
- each position vector instead of the average of each position vector (average of horizontal vector component V x , average of vertical vector component V y ), for example, each position vector It is possible to adopt a value (representative value) obtained by a statistical method from each position vector, such as the maximum or minimum value or standard deviation value. wear.
- a blood vessel image (a blood vessel image after embossing processing and before binarization (blood vessel image) obtained in an intermediate process of blood vessel extraction processing)
- the image portion data DB dj) is adopted.
- the blood vessel and the other parts are clearly distinguished, and the luminance of the blood vessel in the blood vessel image is As shown in Fig. 9, the information represents the actual state of the cross section. However, this information is discarded in the blood vessel image after the blood vessel extraction processing (binary blood vessel image) as shown in FIG. 10, for example, as shown in FIGS. 11 (A) and 11 (B). In addition, even if the cross sections of blood vessels different from each other are represented, the rate of being the same increases through the extraction process.
- the luminance value of the target block ABL in the current image IM 1 from the previous image IM 2 When searching for a process that minimizes the difference (Fig. 8 (B)), many blocks with the same or substantially the same luminance value as the luminance value of the target block ABL will appear. For this reason, it is not possible to search for the block R B L that truly corresponds to the target process A B L. As a result, when the calculation accuracy of the amount of motion decreases, a situation will occur.
- the motion amount calculation unit 25 uses the blood vessel image (the blood vessel image after the embossing process and before the binarization (the blood vessel image) obtained as an object for calculating the motion amount for the blood vessel image. Partial data DB dj)) is adopted.
- the plurality of blocks of interest in the current image IM 1 are generally all the pixels of the current image IM 1, but are designed to be a part of the part constituting the finger or blood vessel in the current image IM 1. Also good.
- the search range for the proxies that minimize the difference from the brightness value in the target block AB L from the previous image IM2 is generally the entire previous image I 2 but the motion detected in the past You may make it the range corresponding to the size of a plurality of attention blocks centering on the position shifted by the amount, and switch the shape of the range according to the amount of time change of the motion amount detected in the past. You may do it.
- the 3D image generation unit 26 uses a 3D space (hereinafter referred to as a “poxel space”) having a cube called a “poxel” as a structural unit as a space to be projected. Define.
- the common part of the silhouette of the finger in the finger image is generated as a three-dimensional finger image (three-dimensional volume) in the poxel space.
- a blood vessel image partial data DB e have DB e 4 is input from the blood vessel extraction unit 24, based on the DB e (n 3 have DB e (n, caught on the blood vessel image
- the common part of the sheet ⁇ in the blood vessel is generated as a three-dimensional blood vessel (three-dimensional volume) in the poxel space.
- the shape extraction unit 27 converts these 3D image data (poxel data) into a 3D volume data DTD. To give.
- the 3D image generator 26 recognizes the viewpoint of each finger image captured from around the finger based on camera information such as focal length and image center, and information on the poxel space, and captures the image from these viewpoints.
- a finger is projected onto the voxel space
- a silhouette area projected onto the voxel space is detected. That is, when the finger image partial data DF di initially input from the image cutout unit 23 is the processing target, the 3D image generation unit 26 uses the finger image based on the finger image partial data DF d as a reference image. For example, as shown in Fig. 13, among the viewpoints around the poxel space, it is placed at the position corresponding to the viewpoint with the rotation angle of 0 [°] and is projected to the farthest from the projection surface of the poxel space. Detect AR F.
- each poxel in the poxel space is back-projected toward the finger image to calculate a projection point, and the projection point of the poxel in the finger image that appears in the finger image
- the silt region is detected by leaving the region as a poxel.
- the 3D image generation unit 26 processes the finger image portion data DF d 3 , DF d 5 s, etc. input after the first finger image portion DF di from the image cutout portion 23.
- the motion amount corresponding to the rotation direction from the reference image to the finger image partial processing target DF d based on the DF d (hereinafter referred to as the rotational motion amount) is used as the motion amount calculation unit. Recognized based on the corresponding movement amount data D FM input from 25.
- the 3D image generator 26 takes the amount of rotational movement as V xl and the value set as the distance from the finger rotation axis to the finger surface as r 1.
- ⁇ roi arc tan (V xl / rl) (1) is used to calculate the rotation angle of the finger image that is the current processing target with respect to the reference image (hereinafter referred to as the first rotation angle) Spci. Judge whether it is less than 360 [°].
- This first rotation angle 0 r If i is less than 360 [°], this means that the entire visual volume (silette area) for multiple finger images captured from the entire circumference of the finger has not yet been detected. . In this case, the three-dimensional image generation unit 26 is immediately before the first rotation angle Spoi and the current processing target. The difference between the finger image in which the visual volume is detected and the rotation angle of the reference image (hereinafter referred to as the second rotation angle) is obtained, and it is determined whether this difference is equal to or greater than a predetermined threshold value.
- the 3D image generation unit 26 processes the finger image partial data DF d input next to the processing target without obtaining the silhouette area of the finger image currently being processed. set to target. As a result, the 3D image generation unit 26 can prevent unnecessary calculation of the silt region.
- the 3D image generation unit 26 has a first rotation angle 0 P with respect to the viewpoint VP s of the reference position IM s , for example, as shown in FIG. It recognizes the viewpoint VP X that forms a in a position corresponding to the convergent point VP X, placing the finger image IM x serving as the current processing target.
- the 3D image generation unit 26 detects, for this finger image IM x , a silhouette region projected from the projection surface of the projection space to the farthest, and then the finger image portion input next to the processing target It is designed to process DF d.
- the 3D image generation unit 26 is configured to display the finger image ⁇ . ⁇ and the view volume immediately before the finger image IM X.
- the finger image IM detected is the amount of motion in the direction orthogonal to the finger rotation direction for u (the vector component in the vertical direction of the finger image currently being processed and the last finger image placed (Average of V y ) is recognized based on the corresponding movement amount data DFM (Fig. 4), and the correction direction (direction parallel to the Z-axis direction of the poxel space) by the amount of movement RD is the viewpoint VP x Correct the position.
- the 3D image generator 2 6 Even if fluctuations occur, the silhouette area can be detected by following the fluctuations. Therefore, compared to the case where the amount of movement in the direction orthogonal to the finger rotation direction is not taken into account, the silhouette area can be accurately determined. It is made to be able to detect.
- the three-dimensional image generation unit 26 determines the finger's shot region in each finger image captured from around the finger until the first rotation angle with respect to the reference image is equal to or greater than 360 ° [°]. Each will be detected.
- the 3D image generation unit 26 designates a common portion of each of the detected silhouette regions. By extracting as a three-dimensional image (three-dimensional volume), for example, as shown in FIG. 15, the finger three-dimensional image is generated.
- the 3D image generator 26 recognizes the viewpoint of each blood tube image captured from around the finger based on the camera information and the information related to the “poxel space” as in the case of the finger stereoscopic image.
- the silhouette regions projected onto the projection space are detected.
- the 3D image generation unit 26 does not detect the projected area projected from the projection surface of the projection space to the innermost part, but is separated from the projection surface in the projection space by a specified length in the depth direction. Detects the blood vessel region projected up to the projection plane. .
- the 3D image generation unit 26 has a binary value based on the blood vessel image partial data DF d 2.
- the blood vessel image is placed at a position corresponding to the viewpoint with a rotation angle of 0 [°] among the viewpoints around the poxel space, and extends in the depth direction from the projection surface in the projection space.
- the detection method of the silt region is the same as that of the finger image.
- the 3D image generation unit 26 processes the blood vessel image partial data DB e 4 , DB e 6 ,... Input from the blood vessel extraction unit 24 after the first blood vessel image data DB e 2.
- the rotational motion amount from the reference image to the binary blood vessel image is recognized based on the corresponding motion amount data DBM input from the motion amount calculation unit 25.
- the 3D image generation unit 26 assumes that the amount of rotational movement is V x 2 and the value set as the distance from the rotation axis of the finger to the blood vessel is r 2.
- ⁇ ro 2 arctan (V x 2 Zr2) (2), the first rotation angle 0 P of the binary blood vessel image with respect to the reference image. 2 is requested. '
- the three-dimensional image generator 26 has the first rotation angle ⁇ r .
- the blood vessels in the binary blood vessel image to be processed are processed in the projection space in the same manner as in the case of the finger three-dimensional image.
- the silhouette area projected in the space from the projection surface to the projection plane separated by a specified length L in the depth direction is detected. .
- the part other than the blood vessel in the living body is not hollow and is blocked by each tissue such as fat, for example, as shown in FIG. 17, in the optical imaging unit 12, the back surface of the imaging surface There is a case where the blood vessel portion existing on the side cannot be copied. Even in this case, the detection target of the silt region is limited to the region projected from the projection surface in the poxel space (projection space) to the projection plane separated by a specified length in the depth direction. As shown in Fig. 8, in the poxel space, the poxel of the projection part (silette area) of the object in each image remains.
- the first rotation angle 0 P with respect to the reference image. 2 is 3 6 0 [°] or more If the silt region is detected until
- the common part (solid line part) is a three-dimensional image of blood vessels that is faithful to the actual object (3 Dimension volume).
- the part of the cylindrical region is the remaining poxel as the non-projected part.
- the three-dimensional image generation unit 26 has a first rotation angle> r with respect to the reference image.
- a common part of each silt region detected so far is extracted as a blood vessel stereoscopic image (three-dimensional volume). As shown in 0., the blood vessel stereoscopic image is generated o
- the shape extraction unit 27 acquires the 3D volume data DTD input from the 3D image generation unit 26, the shape extraction unit 27, for example, as shown in FIG.
- the position of the joint JNL in the finger body image is recognized based on the position data DPi input from the finger joint detection unit 21.
- the shape extraction unit 27 uses the position of the joint JNL as a reference, extracts cross-sectional shape values in a plurality of cross-sections having a predetermined positional relationship with the joint position, and generates each cross-sectional shape value as identification data DIS .
- This identification data D S is registered in the memory 13 in the blood vessel registration mode, and is compared with the identification data registered in the memory 13 in the authentication mode.
- the shape extraction unit 27 has a cross-section SC 1 that passes through the position of the joint and is parallel to the joint, and a direction perpendicular to the joint from the joint position (the length of the finger Cross-sections SC 2 and SC 3 that pass through a position separated by a first distance DS 1 and D 'S 2 in parallel to the joint, and the first from the joint position to the longitudinal direction of the finger Cross sections SC4 and SC5 that pass through positions separated by the second distances DS3 and DS4 that are larger than this distance and are parallel to the joint are determined as the cross-sectional shape value extraction targets.
- the shape extraction unit 27 for each of the cross sections SC 1 to SC 5, the shape extraction unit 27, for example, as shown in FIG. 23, the outer circumference 0 C of the cross section, the area S FA, the central position CP, the major axis MA 1 passing through the central position CP, and The minor axis MA 2 and the position of the blood vessel existing in the cross section (shown by hatching in the figure) are extracted as the cross-sectional shape values. '.
- control unit 10 in the authentication device 1 is the target space.
- the common part of the silhouette of the finger in the finger is generated as a 3D finger image (Fig. 15) from a plurality of finger images around the body part.
- control unit 10 in the target space from a plurality of blood vessel images with a view around the living body part, a common portion of the syrup ⁇ ⁇ ⁇ ⁇ in the blood vessel inside the finger that appears in the image is a blood vessel stereoscopic image ( Fig. 20)
- control unit 10 determines the position of the blood vessel in the cross section having a predetermined positional relationship with respect to the position of the joint JN L (FIG. 21) in the three-dimensional finger image (shaded portion in FIG. 23). Extracted as an identification de overnight.
- the control unit 10 extracts the identification data as representing the shape of a part having a certain relationship with the finger stereoscopic image while representing the shape of the part of the blood vessel stereoscopic image itself.
- Both the image and the blood vessel stereoscopic image can be discretely expressed by the position of the blood vessel.
- the amount of information of the identification target is larger than when only the finger stereoscopic image and the blood vessel stereoscopic image are used as the identification information. It is possible to improve the authentication accuracy while suppressing the above.
- the cross-sectional shape value (blood vessel position) representing the shape in the cross section of the blood vessel stereoscopic image
- the cross sectional shape representing the shape in the cross section of the finger stereoscopic image.
- the status values (peripheral OC, product SFA, center position CP, major axis MA 1 and minor axis MA 2 passing through the center position CP) are also identified.
- the identification data is divided into a finger solid for every five sections SC 1 to SC 5 (FIG. 2 2) having a predetermined positional relationship with respect to the position of the joint JNL (FIG. 21).
- the cross-sectional shape values of both the image and the blood vessel stereoscopic image are used as identification data.
- control unit 10 can express the structure of the finger in detail, and can further improve the authentication accuracy.
- control unit 10 detects a finger joint JNL appearing in the finger image before generating the finger stereoscopic image and the blood vessel stereoscopic image, and expands the joint JNL in the row or column direction of the finger image and the blood vessel image.
- the finger image and the blood vessel image are rotationally corrected so that the angle formed with the direction becomes a predetermined angle.
- the control unit 10 when generating the finger stereoscopic image or the blood vessel stereoscopic image, the control unit 10 can accurately obtain the common part of the silhouette in the finger or blood vessel that appears in the image of the generation base, and as a result, the control unit 10 can further obtain the same. Authentication accuracy can be improved. .
- the rotation correction standard is the same joint JNL as the cross-sectional standard. Therefore, compared to the case where the standard is separate, the finger stereoscopic image or the blood vessel stereoscopic image is generated. It becomes possible to reduce processing load.
- the control unit 10 when generating a finger stereoscopic image and a blood vessel stereoscopic image, the control unit 10 causes the imaging unit 12 to blink the near-infrared light source in the imaging unit 12 to blink the finger image. And blood vessel images are alternately captured. Then, for the finger image and the blood vessel image, the control unit 10 calculates the finger movement amount or the blood vessel movement amount that appears in the image selected as the calculation target and the image input immediately before the image (see FIG. 8 etc.)
- the control unit 10 recognizes the viewpoint of each finger image and each blood vessel image from the amount of movement, and the finger or blood displayed on the image from the viewpoint position of the image.
- the common part of the projection area (Fig. 13 or Fig. 16) projected onto the voxel space is represented by a finger stereo image (Fig. 15) or a blood vessel stereo image (Fig. 20). ).
- control unit 10 can generate the finger standing heel image and the blood vessel stereoscopic image from the image captured by the single imaging unit 12, the stereoscopic image can be generated from the images captured by the plurality of power cameras. Compared to the generation of, it can be downsized and is useful when mounted on a portable terminal device such as a PDA or mobile phone.
- the position of the blood vessel in the cross section having a predetermined positional relationship with the position of the joint J NL (FIG. 21) in the finger stereoscopic image among the finger stereoscopic image and the blood vessel stereoscopic image By extracting the hatched portion in Fig. 23 as the identification data, both the finger 3D image and blood vessel 3D image can be expressed in a dispersive manner depending on the position of the blood vessel.
- An authentication device 1 that can improve the authentication accuracy while suppressing the amount of information to be identified can be realized. ⁇
- the value (cross-sectional shape value) representing the shape of the plurality of cross-sections SC 1 to SC 5 (FIG. 2 2) having a predetermined positional relationship with respect to the reference position is extracted.
- a volume surrounded by a pair of cross-sections selected from the cross-section and the outer shape of the finger may be extracted.
- the pair of cross sections to be selected may be any combination such as cross sections S C 1 and S C 5, or cross sections S C 1 and S C 2, and two or more pairs of cross sections may be selected.
- some of these may be omitted.
- an object to be extracted as a cross-sectional shape value is input or selected via the operation unit 11 (FIG. 1), and the input or selected cross-sectional shape value is extracted. May be. In this way, the content and number of extractions for the cross-sectional shape value can be used as incidental password information that only the user can know, thus further improving the authentication accuracy while reducing the amount of information to be identified. can do.
- the number of extracted cross sections is input or selected via the operation unit 11 (FIG. 1), and the cross sectional shape value corresponding to the input or selected extracted number is extracted. It may be. In this way, since the number of extracted cross sections can be set as incidental password information that only the user can know, the authentication accuracy can be further improved while suppressing the amount of information to be identified.
- Distance of DS 3 and DS 4 The cross-sections SC 4 and SC 5 that pass through the separated positions and are parallel to the joint are described, but the present invention is not limited to this, and other positional relationships may be used.
- all or a part of the cross-sections SC1 to SC5 may be changed to a cross-section that forms a predetermined angle with respect to a plane parallel to the joint, and the reference position is changed to the position of the joint.
- the reference position is appropriately changed according to the type of the image of the living body part to be adopted, such as a lifeline.
- the positional relationship may be prepared as a plurality of patterns, and a cross section having the positional relationship of the selected pattern among these patterns may be extracted with respect to the reference position.
- a blood vessel image is applied as a plurality of second images with a view around the living body part.
- the present invention is not limited thereto, and for example, a nerve or the like is applied. Can do.
- the second image is acquired from an optical camera (imaging unit 12 (FIG. 1)) mounted on the authentication device 1, but as a device separate from the authentication device 1,
- a blood vessel mixed with a contrast agent may be acquired from an imaging device that images with X-rays.
- the near-infrared light source in the imaging unit 12 is turned on for a predetermined lighting period at predetermined intervals.
- the present invention is not limited to this.
- an optical film that transmits only visible light is arranged at predetermined intervals at predetermined intervals. You may make it arrange
- the blood vessel registration mode and the authentication mode are executed based on the program stored in the ROM.
- the present invention is not limited to this, and the CD Register blood vessels based on programs acquired by installing from a program storage medium such as (Compact Disc), DVD (Digital Versatile Disc), or semiconductor memory, or by downloading from a program providing server on the Internet.
- the mode and authentication mode may be executed.
- control unit 10 executes the registration process and the authentication process has been described.
- the present invention is not limited to this, and part of these processes is performed by the graphics workstation. It may be executed at
- the authentication device 1 having the imaging function, the verification function, and the registration function has been described.
- the present invention is not limited to this, and the function is determined according to the application. Alternatively, a part of each function may be applied to a single device. Industrial applicability.
- the present invention can be used in the field of biometric authentication.
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Vascular Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Collating Specific Patterns (AREA)
- Image Input (AREA)
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/528,540 US9031315B2 (en) | 2007-02-26 | 2008-02-25 | Information extraction method, information extraction device, program, registration device, and verification device |
EP08721134A EP2131324A1 (en) | 2007-02-26 | 2008-02-25 | Information extracting method, information extracting device, program, registering device and collating device |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2007046091A JP5061645B2 (ja) | 2007-02-26 | 2007-02-26 | 情報抽出方法、情報抽出装置、プログラム、登録装置及び照合装置 |
JP2007-046091 | 2007-02-26 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2008105548A1 true WO2008105548A1 (ja) | 2008-09-04 |
Family
ID=39721370
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2008/053714 WO2008105548A1 (ja) | 2007-02-26 | 2008-02-25 | 情報抽出方法、情報抽出装置、プログラム、登録装置及び照合装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US9031315B2 (ja) |
EP (1) | EP2131324A1 (ja) |
JP (1) | JP5061645B2 (ja) |
KR (1) | KR20090115739A (ja) |
CN (1) | CN101622642A (ja) |
WO (1) | WO2008105548A1 (ja) |
Families Citing this family (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110007951A1 (en) * | 2009-05-11 | 2011-01-13 | University Of Massachusetts Lowell | System and method for identification of fingerprints and mapping of blood vessels in a finger |
EP2546798A4 (en) * | 2010-03-10 | 2017-08-16 | Fujitsu Limited | Biometric authentication device and biometric authentication method |
US9030536B2 (en) | 2010-06-04 | 2015-05-12 | At&T Intellectual Property I, Lp | Apparatus and method for presenting media content |
US9787974B2 (en) | 2010-06-30 | 2017-10-10 | At&T Intellectual Property I, L.P. | Method and apparatus for delivering media content |
US8918831B2 (en) | 2010-07-06 | 2014-12-23 | At&T Intellectual Property I, Lp | Method and apparatus for managing a presentation of media content |
US9049426B2 (en) | 2010-07-07 | 2015-06-02 | At&T Intellectual Property I, Lp | Apparatus and method for distributing three dimensional media content |
US9232274B2 (en) | 2010-07-20 | 2016-01-05 | At&T Intellectual Property I, L.P. | Apparatus for adapting a presentation of media content to a requesting device |
US9032470B2 (en) | 2010-07-20 | 2015-05-12 | At&T Intellectual Property I, Lp | Apparatus for adapting a presentation of media content according to a position of a viewing apparatus |
US9560406B2 (en) | 2010-07-20 | 2017-01-31 | At&T Intellectual Property I, L.P. | Method and apparatus for adapting a presentation of media content |
US8994716B2 (en) * | 2010-08-02 | 2015-03-31 | At&T Intellectual Property I, Lp | Apparatus and method for providing media content |
US8438502B2 (en) | 2010-08-25 | 2013-05-07 | At&T Intellectual Property I, L.P. | Apparatus for controlling three-dimensional images |
US8947511B2 (en) | 2010-10-01 | 2015-02-03 | At&T Intellectual Property I, L.P. | Apparatus and method for presenting three-dimensional media content |
JP5809792B2 (ja) * | 2010-11-04 | 2015-11-11 | 株式会社日立製作所 | 生体認証装置および方法 |
US8655027B1 (en) * | 2011-03-25 | 2014-02-18 | The United States of America, as represented by the Director, National Security Agency | Method of image-based user authentication |
US9445046B2 (en) | 2011-06-24 | 2016-09-13 | At&T Intellectual Property I, L.P. | Apparatus and method for presenting media content with telepresence |
US8947497B2 (en) | 2011-06-24 | 2015-02-03 | At&T Intellectual Property I, Lp | Apparatus and method for managing telepresence sessions |
US9030522B2 (en) | 2011-06-24 | 2015-05-12 | At&T Intellectual Property I, Lp | Apparatus and method for providing media content |
US9602766B2 (en) | 2011-06-24 | 2017-03-21 | At&T Intellectual Property I, L.P. | Apparatus and method for presenting three dimensional objects with telepresence |
US8587635B2 (en) | 2011-07-15 | 2013-11-19 | At&T Intellectual Property I, L.P. | Apparatus and method for providing media services with telepresence |
US8625873B2 (en) * | 2012-02-24 | 2014-01-07 | Kabushiki Kaisha Toshiba | Medical image processing apparatus |
KR20130109817A (ko) * | 2012-03-28 | 2013-10-08 | 삼성전자주식회사 | 손가락 패턴을 이용한 손 모양 인식 장치 및 방법 |
KR101572487B1 (ko) * | 2013-08-13 | 2015-12-02 | 한국과학기술연구원 | 환자와 3차원 의료영상의 비침습 정합 시스템 및 방법 |
WO2016042039A1 (en) * | 2014-09-16 | 2016-03-24 | Foundation For Research And Technology - Hellas (Forth) | Gesture recognition apparatuses, methods and systems for human-machine interaction |
US20160162673A1 (en) * | 2014-12-05 | 2016-06-09 | Gershom Kutliroff | Technologies for learning body part geometry for use in biometric authentication |
TWI610292B (zh) * | 2015-11-19 | 2018-01-01 | 瑞鼎科技股份有限公司 | 驅動電路及其運作方法 |
KR101626837B1 (ko) * | 2015-12-08 | 2016-06-02 | 주식회사 나르테크놀로지 | 손가락 마디 및 지정맥 기반의 융합형 생체 인증 방법 및 그 장치 |
JP6765057B2 (ja) * | 2016-03-18 | 2020-10-07 | パナソニックIpマネジメント株式会社 | 画像生成装置、画像生成方法およびプログラム |
WO2018063914A1 (en) * | 2016-09-29 | 2018-04-05 | Animantis, Llc | Methods and apparatus for assessing immune system activity and therapeutic efficacy |
CN108764127A (zh) * | 2018-05-25 | 2018-11-06 | 京东方科技集团股份有限公司 | 纹理识别方法及其装置 |
CN110381265B (zh) * | 2019-07-12 | 2021-08-31 | 武汉恒新动力科技有限公司 | 筒状工件内壁图像获取方法、设备及计算机可读存储介质 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002175529A (ja) | 2000-12-06 | 2002-06-21 | Matsushita Electric Ind Co Ltd | 個人識別装置 |
JP2003067726A (ja) * | 2001-08-27 | 2003-03-07 | Sanyo Electric Co Ltd | 立体モデル生成装置及び方法 |
JP2003093369A (ja) * | 2001-09-21 | 2003-04-02 | Sony Corp | 認証処理システム、および認証処理方法、並びにコンピュータ・プログラム |
JP2007000219A (ja) * | 2005-06-22 | 2007-01-11 | Hitachi Ltd | 個人認証装置 |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10307919A (ja) * | 1997-05-01 | 1998-11-17 | Omron Corp | 本人特定装置 |
US6496594B1 (en) * | 1998-10-22 | 2002-12-17 | Francine J. Prokoski | Method and apparatus for aligning and comparing images of the face and body from different imagers |
JP2002092616A (ja) * | 2000-09-20 | 2002-03-29 | Hitachi Ltd | 個人認証装置 |
JP2003070021A (ja) * | 2001-08-27 | 2003-03-07 | Sanyo Electric Co Ltd | ポータブル三次元データ入力装置及び立体モデル生成装置 |
JP2004070792A (ja) | 2002-08-08 | 2004-03-04 | Telecommunication Advancement Organization Of Japan | ボクセルデータ符号化方式 |
US8190239B2 (en) * | 2002-09-03 | 2012-05-29 | Fujitsu Limited | Individual identification device |
JP4207717B2 (ja) * | 2003-08-26 | 2009-01-14 | 株式会社日立製作所 | 個人認証装置 |
JP4671027B2 (ja) * | 2005-05-02 | 2011-04-13 | ソニー株式会社 | 認証装置、認証方法及びプログラム |
US20070286462A1 (en) * | 2006-04-28 | 2007-12-13 | David Usher | System and method for biometric retinal identification |
-
2007
- 2007-02-26 JP JP2007046091A patent/JP5061645B2/ja not_active Expired - Fee Related
-
2008
- 2008-02-25 EP EP08721134A patent/EP2131324A1/en not_active Withdrawn
- 2008-02-25 CN CN200880006205A patent/CN101622642A/zh active Pending
- 2008-02-25 KR KR1020097017660A patent/KR20090115739A/ko not_active Application Discontinuation
- 2008-02-25 WO PCT/JP2008/053714 patent/WO2008105548A1/ja active Application Filing
- 2008-02-25 US US12/528,540 patent/US9031315B2/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002175529A (ja) | 2000-12-06 | 2002-06-21 | Matsushita Electric Ind Co Ltd | 個人識別装置 |
JP2003067726A (ja) * | 2001-08-27 | 2003-03-07 | Sanyo Electric Co Ltd | 立体モデル生成装置及び方法 |
JP2003093369A (ja) * | 2001-09-21 | 2003-04-02 | Sony Corp | 認証処理システム、および認証処理方法、並びにコンピュータ・プログラム |
JP2007000219A (ja) * | 2005-06-22 | 2007-01-11 | Hitachi Ltd | 個人認証装置 |
Also Published As
Publication number | Publication date |
---|---|
US9031315B2 (en) | 2015-05-12 |
EP2131324A1 (en) | 2009-12-09 |
JP5061645B2 (ja) | 2012-10-31 |
US20100098299A1 (en) | 2010-04-22 |
JP2008210141A (ja) | 2008-09-11 |
CN101622642A (zh) | 2010-01-06 |
KR20090115739A (ko) | 2009-11-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2008105548A1 (ja) | 情報抽出方法、情報抽出装置、プログラム、登録装置及び照合装置 | |
CN108875524B (zh) | 视线估计方法、装置、***和存储介质 | |
CN106407914B (zh) | 用于检测人脸的方法、装置和远程柜员机*** | |
CN109740491B (zh) | 一种人眼视线识别方法、装置、***及存储介质 | |
US8406484B2 (en) | Facial recognition apparatus, method and computer-readable medium | |
JP6165540B2 (ja) | 血管画像撮影装置及び端末 | |
CN106547341B (zh) | 注视***及其跟踪注视的方法 | |
US20150347833A1 (en) | Noncontact Biometrics with Small Footprint | |
EP2339507B1 (en) | Head detection and localisation method | |
US7760916B2 (en) | Registration apparatus, collation apparatus, image correction method and program | |
JP2007136000A (ja) | 視線検出装置、視線検出方法、および視線検出プログラム | |
US20220148218A1 (en) | System and method for eye tracking | |
CN110505398B (zh) | 一种图像处理方法、装置、电子设备及存储介质 | |
JP6846330B2 (ja) | 生体認証装置および生体認証システム | |
JP2010170348A (ja) | 画像処理装置、生体認証装置、画像処理方法及びプログラム | |
JP2007041831A (ja) | 生体判別装置および認証装置ならびに生体判別方法 | |
WO2008105545A1 (ja) | 情報抽出方法、登録装置、照合装置及びプログラム | |
Benalcazar et al. | A 3D iris scanner from multiple 2D visible light images | |
JP2008203995A (ja) | 物体形状生成方法、物体形状生成装置及びプログラム | |
JP2011258209A5 (ja) | 手の甲の認証端末及び手の甲の認証方法 | |
JP2004126738A (ja) | 3次元計測を用いた個人認証装置および認証方法 | |
CN111353388A (zh) | 活体检测方法、装置、电子设备及存储介质 | |
JP4636338B2 (ja) | 表面抽出方法、表面抽出装置及びプログラム | |
Zonneveld | 3D finger vein patterns: acquisition, reconstruction and recognition | |
Nikisins et al. | Increasing the reliability of biometric verification by using 3D face information and palm vein patterns |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200880006205.3 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 08721134 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2008721134 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 4964/CHENP/2009 Country of ref document: IN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 12528540 Country of ref document: US Ref document number: 1020097017660 Country of ref document: KR |
|
NENP | Non-entry into the national phase |
Ref country code: DE |