JPH08161495A - Fingerprint collation device and fingerprint marking discrimination method - Google Patents

Abstract

PURPOSE: To ensure discrimination of a fingerprint marking state with respect to the fingerprint collation device and fingerprint marking method. CONSTITUTION: The device includes a fingerprint image pickup section 1 picking up an image through an incident reflected light from a light emitted from a light source, an in-block average luminance calculation section 2 dividing a discrimination area of a fingerprint into plural blocks and calculating an average luminance in a block based on a multi-value image signal entered from the fingerprint image pickup section 1 via an interface section 6, a luminance deviation calculation arrangement processing section 3 obtaining a deviation between the in-lock luminance and a luminance of each picture element in a discrimination area to obtain number of picture elements corresponding to the deviation, a marking discrimination section 4 discriminating the fingerprint image pickup section 1 reaches a fingerprint marking state when number of picture elements within a prescribed range in the vicinity of deviation zero as to number of the picture elements corresponding to the deviation does not exceed a threshold number and discriminating the fingerprint image pickup section 1 is not in a fingerprint marking state when number of picture elements within a prescribed range in the vicinity of deviation zero exceeds the threshold number, and a fingerprint collation section 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fingerprint collation device and a fingerprint imprint determination method which prevent erroneous determination in fingerprint collation. There is known a fingerprint collation device that registers a fingerprint in advance and images the fingerprint with a fingerprint imaging unit when verifying the identity and collates it with a registered fingerprint. In that case, whether or not the fingerprint to be collated is correctly imaged. It is necessary to judge whether

[0002]

2. Description of the Related Art FIG.
Is a camera, 22 is a light source, 23 is a prism, 24 is an elastic film, and 25 is a finger for imprinting a fingerprint. The camera 21 can be, for example, a CCD camera, and is arranged at an angle equal to or less than the critical angle θ with respect to the imprint surface (surface of the elastic film 24), and when the imprint surface is illuminated from the light source 22 via the prism 23. ,
When the finger 25 is not placed, the reflected light does not enter the camera 21, and when the finger 25 is placed on the elastic film 24 of the imprinting surface, the ridge line and the valley line forming the fingerprint of the finger 25 are formed. The reflected light corresponding to is incident on the camera 21,
A signal obtained by imaging the fingerprint is output from 1.

Further, the elastic film 24 is for improving the adhesion to the finger 25 so that a clear fingerprint can be picked up even with a dry finger, and a certain degree of abrasion resistance is required. is there.
Therefore, there is no colorless and transparent material that can be used. Therefore, when strong light is incident on the embossing surface from the outside, the elastic film 2
4 may generate scattered light and enter the camera 21. When the elastic film 24 is flawed or dust is attached, the light emitted from the light source 22 causes reflected light to enter the camera 21.

In the fingerprint collation device, a finger 25 is placed on the fingerprint imprinting surface (on the elastic film 24) and the camera 21 picks up and processes a fingerprint. The image signal from the camera 21 is taken as a fingerprint. It is necessary to judge whether it is due to. Therefore, in the conventional fingerprint collation device,
For example, (a) when a finger is placed on the embossing surface, the reflected light from the finger is incident on the camera 21. Therefore, the average brightness in the determination area used for fingerprint collation in the image captured by the camera 21 is calculated. When the average brightness exceeds a preset threshold value, it is estimated that the reflected light from the finger 25 is incident on the camera 21, and it is determined that the fingerprint has been imprinted. A method of performing fingerprint collation is known.

(B) A method is also known in which the judgment area is divided into a plurality of blocks, and when the number of blocks in which the average luminance in each block exceeds a threshold value is equal to or more than a predetermined value, an impressed state is determined. . If the number of blocks in the determination area is j, the number of pixels in the block is m, and the luminance of pixels in the determination area is P _qn , the average luminance A _q of a block is A _q = (Σ _{n = 1} ^{n =} It can be expressed as ^m P _qn ) / m (1). Then, for a certain threshold value a, the total number of blocks whose average brightness is A _q > a is calculated to obtain N
There is also known a method in which when the total number N has a relation of N> b with respect to a certain threshold value b, fingerprint imprinting is determined.

[0006]

SUMMARY OF THE INVENTION (a) of the above-mentioned conventional example.
When the average brightness in the determination area of ∘ exceeds the threshold, the method of determining fingerprint imprinting is such that when strong light is incident on the fingerprint imprinting surface such as direct sunlight, the scattered light of the elastic film 24 is reflected by the camera 21. The entire screen becomes a white screen and the average brightness rises. That is, an error will occur in the imprint determination.

Further, in the conventional example (b), the number N in which the average brightness of each block in the determination area exceeds the threshold value a is determined, and when the number N exceeds the threshold value b, fingerprint imprinting is determined by a light source. Even if it is possible to relieve the uneven illumination of No. 22, (a) of the conventional example
Similarly to the case, when strong light is incident on the imprinting surface, the average brightness of each block becomes large, so that an error occurs in impressing determination. It is an object of the present invention to provide a fingerprint collation device and a fingerprint imprint determination method that prevent such an improper imprint determination.

[0008]

A fingerprint collation device and a fingerprint imprint determination method according to the present invention include a fingerprint image pickup unit 1 configured to enter reflected light by illumination from a light source and pick up an image. The block average brightness calculation unit 2 calculates the average brightness in the block based on the multi-valued image signal from the fingerprint imaging unit 1, and the block average brightness from the block average brightness calculation unit 2. And a luminance deviation of each pixel in the determination area, and the number of pixels corresponding to the deviation is calculated. A luminance deviation calculation / arrangement unit 3 and a pixel corresponding to the luminance deviation calculation / deviation from the arrangement unit 3 Regarding the number, when the number of pixels within a predetermined range where the deviation is close to zero does not exceed the threshold value, the imprint determination unit 4 that determines that the fingerprint imaging unit 1 is in the fingerprint imprinting state, and the fingerprint by the imprint determination unit 4 Determine the pressed state And a fingerprint verification unit 5 for fingerprint identification based on multi-valued image signal from the fingerprint image pickup unit 1 when.

Further, the fingerprint imprinting determination method in the fingerprint collating apparatus is such that the average brightness is obtained for each of a plurality of blocks in the fingerprint determination area based on the multi-valued image signal picked up by the fingerprint image pickup unit 1, and each block is determined. The difference between the average brightness and the brightness of the pixels in the block is calculated, and the number of pixels corresponding to the deviation is calculated for the entire judgment area. When the horizontal axis is the deviation and the vertical axis is the number of pixels, the deviation is zero. To determine whether or not the number of pixels within a predetermined range centered on a point shifted by a predetermined value in the horizontal axis exceeds a threshold value, and if not, determine that the fingerprint imprinting state is set. It is a waste.

[0010]

The multi-valued image signal from the fingerprint image pickup unit 1 including a CCD camera or the like is applied to the intra-block average brightness calculation unit 2 via the interface unit 6, and the intra-block average brightness calculation unit 2 The determination area is divided into a plurality of blocks, and the average brightness in each block is calculated. The luminance deviation calculation / arrangement unit 3 obtains the deviation between the average luminance in the block and the luminance of each pixel in the block, processes all the pixels in the determination area, and finds the correspondence between the deviation and the number of pixels. The imprint determination unit 4 determines that external light has entered the fingerprint imaging unit 1 when the number of pixels within a predetermined range in which the deviation is near zero exceeds a threshold value, and when the number of pixels does not exceed the threshold, the fingerprint imaging unit 1 It is determined that the fingerprint imprinting state is set, and the multivalued image signal from the fingerprint image pickup unit 1 at that time is binarized in the binarization unit 7, and is registered in the storage unit 8. The fingerprint is collated by the fingerprint collation unit 5, and the collation judgment is output.

Further, the imprinting determination method obtains the average luminance for each of a plurality of blocks in the fingerprint determination area, obtains the deviation between the luminance of each pixel of this block and the average luminance within the block, and then determines the entire determination area. Then, the relationship between the deviation and the number of pixels is calculated. When the horizontal axis represents the deviation and the vertical axis represents the number of pixels, when external light is incident, or conversely, when the incident light on the fingerprint image pickup unit 1 is only a slight reflected light, most of the pixels are in the block. Deviation from the average brightness is zero or close to it,
The number of pixels near the deviation of zero becomes large enough to exceed the threshold value. On the other hand, in the fingerprint imprinting state, the reflected light corresponding to the fingerprint is incident on the fingerprint image pickup unit 1, so that the number of pixels in the vicinity of the deviation of zero decreases, and the bimodal characteristic of the bimodal characteristics is centered around the deviation of zero. The number of pixels becomes a distribution, and the number of pixels near the deviation of zero does not exceed the threshold value. In addition, the pixel number distribution characteristic with respect to the deviation does not have symmetry centered on the deviation zero in correspondence with the electrical characteristics of the fingerprint image pickup unit 1 and the configuration of each part, so that the horizontal axis direction from the point of the deviation zero. It is determined whether or not the number of pixels within a predetermined range exceeds a threshold value, centering on a point obtained by shifting a predetermined value.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is an explanatory diagram of a fingerprint collation apparatus according to an embodiment of the present invention, in which 1 is a fingerprint image pickup unit, 2 is an average brightness calculation unit in a block, 3 is a brightness deviation calculation unit, 4 is an imprinting unit. Judgment part,
Reference numeral 5 is a fingerprint collation unit, 6 is an interface unit, 7 is a binarization unit, and 8 is a storage unit. As shown in FIG. 5, the fingerprint image pickup unit 1 includes a camera 21, a light source 22, a prism 23, and an elastic film 2.
4 is included.

The multi-valued image signal from the fingerprint image pickup unit 1 is
For example, it is converted into a digital signal in the interface unit 6 and added to the intra-block average luminance calculation unit 2.
The in-block average brightness calculating unit 2, the brightness deviation calculating unit, the arraying unit 3, and the imprint determining unit 4 can be realized by logic circuits having respective functions, but can be easily realized by an arithmetic processing function of a microprocessor or the like. Can be realized.

FIG. 2 is a flow chart of an embodiment of the present invention, which shows an operation of fingerprint imprinting determination by steps (a) to (f), which will be described below with reference to FIG. The picked-up multi-valued image picked up by the fingerprint pick-up unit 1 is taken in via the interface unit 6 (a), and the intra-block average brightness calculation unit 2 calculates the intra-block average brightness in each determination area (b).

For example, the imaging area 11 shown in FIG.
If a 2 × 512 pixel configuration is used, the determination area 12
Can have, for example, a 400 × 320 pixel configuration, and the determination area 12 is divided into a plurality of blocks 13. Each block 13 can have, for example, a 16 × 16 pixel configuration.

Similarly to the above, if the number of blocks 13 in the judgment area 12 is j, the number of pixels in the block is m, and the brightness of the pixels in the judgment area 12 is P _qn , the average brightness A _q in the block is It can be obtained by the equation (1). As described above, the number of pixels in the determination area 12 is 400 × 32
0 and the number of pixels of the block 13 is 16 × 16, j =
It becomes 500 and m = 256.

The average luminance within the block is obtained (b), and further,
The difference between the brightness of each pixel and the average brightness in each block is calculated, and the number of pixels having the same difference is calculated (c). That is, in the brightness deviation calculation / arrangement unit 3, the average brightness A _q in the block, the brightness P _qn of each pixel in the block and the deviation are obtained, and the deviations are obtained for all the pixels in the determination area 12, Find the number of pixels with the same deviation.

Then, the pixel number distribution with respect to the deviation is obtained so that the horizontal axis represents the deviation and the vertical axis represents the pixel number. That is, the luminance deviation calculation / arrangement unit 3 accumulates the number of pixels having the same deviation and arranges the number of pixels corresponding to the deviation using a table or the like. FIG. 4 shows an example of the distribution of the number of pixels with respect to the deviation. In FIG. 4, a curve A shows the case where external light is incident, and a curve B shows the case of fingerprint imprinting. That is, when external light enters, the blocks 13 in the determination area 12 have almost the same brightness,
The number of pixels near the deviation of 0 increases. Therefore, it has a single peak characteristic. Ideally, if the determination area 12 has uniform brightness, the number of pixels with a deviation of 0 is 128000 in the above case, and the number of pixels with other deviations is 0. As described above, when the fingerprint imprinting state is not set, when the threshold value is σ, since the single-peak characteristic is obtained, the threshold value σ is exceeded.

When a fingerprint is picked up, the luminance distribution corresponds to a minute pattern of ridges and valleys of the fingerprint, so the number of pixels in the vicinity of the deviation of 0 becomes relatively small. That is, the number of pixels having the same brightness as the average brightness in the block is reduced, and the distribution of the number of pixels corresponding to the deviation is obtained over a relatively wide range of the deviation, and the number of pixels is maximum at a certain value of deviation + or deviation −. It becomes a characteristic. In this case, the maximum number of pixels does not exceed the threshold σ.

In practice, the relationship between the deviation and the number of pixels depends on the maximum value of the curve A or the maximum value of the curve A depending on the characteristics of the fingerprint image pickup section 1, the amplification characteristics of the interface section 6 and the like, the digitization characteristics of the multivalued image signal, and the like. The valley portion of the curve B does not coincide with the position where the deviation is 0. Therefore, a range of ± β is defined around the point shifted in the horizontal axis direction by a predetermined value α corresponding to the characteristic from the position of the deviation 0, and it is checked whether the number of pixels in this range exceeds the threshold σ ( Step (d) of FIG. 2. Although FIG. 4 shows a case where the position is shifted from the position where the deviation is 0 in the −deviation direction, it may be shifted in the + deviation direction depending on the characteristics of each part. The value of ± β is selected in consideration of the shapes of the curves A and B, and includes the range in which the number of pixels near the deviation 0 of the curve B is small, and the number of pixels near the deviation 0 of the curve A Setting can be made to include a large range to reduce judgment errors.

Then, it is judged whether or not there is a pixel number exceeding the threshold value σ within the range of ± β (step (e) in FIG. 2),
If the number of pixels exceeds the threshold σ, it is not due to the fingerprint imprinting state but due to external light, etc., so proceed to step (a), and if there is no number of pixels exceeding the threshold σ, it is judged as fingerprint imprinting. , Fingerprint registration and collation processing are performed (f). That is, when the imprint determination unit 4 performs the process of step (e) and determines that the fingerprint is imprinted, the binarization unit 7 binarizes the fingerprint and registers it in the storage unit 8 when the fingerprint is registered. At the time of collation, the registered fingerprint is read from the storage unit 8 and collated by the fingerprint collation unit 5 to obtain a collation determination output indicating whether or not the person is confirmed.

[0022]

As described above, the fingerprint collation device according to the present invention includes the fingerprint image pickup section 1 and the average brightness calculation section 2 in the block.
A luminance deviation calculation / arrangement unit 3, an imprint determination unit 4, and a fingerprint collation unit 5, and obtains a deviation between the average luminance in each block of the determination area and the luminance of each pixel to obtain pixels with the same deviation. The number of pixels is determined, and it is determined whether or not the number of pixels exceeds the threshold value σ.If the number of pixels does not exceed the threshold value σ, it is determined that the fingerprint is being imprinted, and the fingerprint collation is performed. When incident, or conversely, when nothing is incident, the number of pixels in the vicinity of the deviation 0 increases, which has the advantage of being able to reliably determine the fingerprint imprinting state.
Even when the fingerprint is registered, it is possible to determine whether or not the fingerprint is correctly read.

Further, the fingerprint image pickup unit 1, the in-block average luminance calculation unit 2, the luminance deviation calculation / arrangement unit 3, and the imprint determination unit 4
In a fingerprint collation device including and, the deviation between the average brightness of each of a plurality of blocks in the determination area and the brightness of the pixels in the block is obtained, the horizontal axis shows the deviation, and the vertical axis shows the number of pixels with the same deviation. Occasionally, when the number of pixels within the range of ± β does not exceed the threshold value σ around the point where the deviation α is shifted by the predetermined value α in the horizontal axis direction, the fingerprint imprinting state is determined. By selecting the shift value α and the determination range β in accordance with the characteristic of 1, there is an advantage that the fingerprint imprinting state can be correctly determined.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a fingerprint matching device according to an embodiment of the present invention.

FIG. 2 is a flowchart of an embodiment of the present invention.

FIG. 3 is an explanatory diagram of a determination area.

FIG. 4 is an explanatory diagram of imprint determination according to the embodiment of the present invention.

FIG. 5 is an explanatory diagram of a fingerprint image pickup unit.

[Explanation of symbols]

 1 Fingerprint imaging unit 2 In-block average luminance calculation unit 3 Luminance deviation calculation and arrangement unit 4 Impression determination unit 5 Fingerprint collation unit 6 Interface unit 7 Binarization unit 8 Storage unit

Claims (2)

[Claims] 1. A fingerprint image pickup unit configured to enter reflected light from a light source to make an image thereof, and a fingerprint determination area is divided into a plurality of blocks, and a multivalued image signal from the fingerprint image pickup unit is used as a basis. In the block average brightness calculation unit for calculating the average brightness in the block, the deviation between the block average brightness from the block average brightness calculation unit and the brightness of each pixel in the determination area The luminance deviation calculation / arrangement unit that obtains the number of pixels corresponding to, and the number of pixels corresponding to the deviation from the luminance deviation calculation / arrangement unit is within a predetermined range in which the deviation is close to zero. When there is not, a fingerprint based on the multi-valued image signal from the fingerprint determining unit that determines that the fingerprint image capturing unit is in the fingerprint impression state and the fingerprint impression state by the fingerprint determining unit when the fingerprint impression state is determined. Fingerprint illumination for verification A fingerprint collation device comprising a joint part. 2. An average luminance is obtained for each of a plurality of blocks in a fingerprint determination area based on a multi-valued image signal picked up by a fingerprint image pickup unit, and the average luminance of each block and the luminance of pixels in the block are calculated. When the number of pixels corresponding to the deviation is obtained for the entire determination area and the horizontal axis is the deviation and the vertical axis is the number of pixels, a predetermined value is set in the horizontal direction from the point where the deviation is zero. Fingerprint collation characterized by including a process of determining whether or not the number of pixels within a predetermined range centered on the point shifted by the Fingerprint impression determination method in device.