JP2508233B2 - Fingerprint registration device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a fingerprint collation system for a keyless entry system, an entrance management system such as a computer room, and more particularly to a fingerprint collation system suitable for use in the fingerprint collation system. Fingerprint registration device.

(Prior Art) Conventionally, in a fingerprint collation system of this type, prior to fingerprint collation, for example, as shown in Japanese Patent Laid-Open No. 62-78286, a characteristic small area is selected from the entire reference fingerprint image. There is one in which a plurality of images are extracted and registered as partial images.

(Problems to be Solved by the Invention) However, in such a configuration, when the quality of each registered partial image is deteriorated due to crushing, blurring, etc., and the minutia that is the characteristic of the fingerprint is not clear, There is a problem that the fingerprint matching accuracy becomes low. Further, when the number of minutiae for identifying each registered partial image is insufficient, there is also a problem that it is difficult to ensure proper matching accuracy.

In view of the above, the present invention aims to provide a fingerprint registration device that creates and registers a fingerprint registration image so that fingerprint matching can always be properly performed.

(Means for Solving the Problems) In solving the above problems, in the invention described in the claims, as shown in FIG. 1, an optical reading unit 1 for optically reading a human fingerprint as a fingerprint image is provided. , A binarizing means 2 for binarizing the fingerprint image into a binarized image; an extracting means 3 for extracting each minutia in an image central region of the binarized image; The initial classifying means 4 for classifying in order from the center of the digitized image, and the number of minutiae belonging to each of a plurality of predetermined ranges in the image central region centering on each classifying minutia by this initial classifying means 4 are checked, Reclassifying means 5 for reclassifying each of the classification minutiaes in descending order of belonging minutiae, and each of the reclassifying minutiae by the reclassifying means 5 among the plurality of predetermined ranges, the belonging minutiae A predetermined range having the largest number of newuchas and a part of another predetermined range having a smaller number of associated minutia than the predetermined range are reclassified minutiae of the one predetermined range and reclassification minutiae of the other predetermined range. Are determined as the first and second window regions so as to eliminate the overlap between the part within the one predetermined range and the part within the other predetermined range, respectively. There is provided a fingerprint registration device provided with a decision registration means 6 for registering two window areas.

(Effects) According to the present invention thus configured, when a human fingerprint is optically read as a fingerprint image by the optical reading unit 1, the fingerprint image is converted into a binary image by the binarizing unit 2. Binarization is performed, and each minutia in the image central region of this binarized image is extracted by the extraction means 3. Then, when these minutiae are classified by the initial classification means 4 in the order of being closer to the center of the binarized image, the re-classification means 5 divides the plurality of minutiae by the initial classification means 4 into a plurality of image center areas in the image central region. The number of minutiae belonging to each of the predetermined ranges is checked, and each classified minutiae is reclassified in descending order of the number of associated minutiae.

Then, the decision registration means 6 causes one predetermined range having the largest number of belonging minutiae among a plurality of predetermined ranges among the plurality of predetermined ranges of the re-sorting minutia and another predetermined range having a smaller number of belonging minutiae than the predetermined range. And a part of
The first and second re-sorting minutiae of one predetermined range and the re-sorting minutiae of another predetermined range are centered so as to eliminate the overlap between a part within one predetermined range and a part of another predetermined range. It is determined as a window area, and these first and second window areas are registered.

As a result, the first window area is specified by the image portion including the larger number of the associated minutiae, which is the center portion where the fingerprint is less distorted, and the second window area is formed next to the first window area and the associated minutiae. It is specified by an image portion which includes a large number and does not overlap a part of one predetermined range and a part of another predetermined range.

As a result, even if the fingerprint image is of low quality, it is possible to provide the registered fingerprint information that can accurately identify the characteristics of the fingerprint with the first and second window areas. Further, by using the registered fingerprint information specified by such first and second window areas, even if the quality of the fingerprint image at the time of matching is low, fingerprint matching can always be performed with high accuracy.

(Embodiment) Hereinafter, one embodiment of the present invention will be described with reference to the drawings. FIG. 2 shows a fingerprint collation system applied to a door lock mechanism 10 provided in a door of a special room such as a computer room that requires entry management. It shows an example to which the present invention is applied. The fingerprint collation system includes a registration button switch 20, which is operated at the time of fingerprint registration to generate a registration operation signal. Match button switch 30
Is operated during fingerprint verification to generate a verification operation signal.
The ten-key pad 40 is operated when the operator designates his / her registered fingerprint image information from among a plurality of registered fingerprint image information registered in the microcomputer 60, as will be described later, and generates a designation code signal. The designated code, which is the content of the designated code signal, corresponds to the fingerprint of the operator known only to the operator.

The image input device 50 optically reads the fingerprint of the operator's finger under the pressing of the operator's finger on the reading screen of the built-in image sensor, and the read result is the fingerprint showing the fingerprint image information. It is generated as an image signal. The microcomputer 60 is constantly supplied with power from the DC power source B and executes a computer program in cooperation with the registration button switch 20, the collation button switch 30, the ten-key pad 40 and the image input device 50 according to the flowchart shown in FIG. During this execution, arithmetic processing necessary for registration and verification of the fingerprint of the operator's finger and drive control of the drive circuit 70 for the door lock mechanism 10 is performed. The above computer program is
It is stored in advance in the ROM of the microcomputer 60.

In the present embodiment configured as described above, the microcomputer 60 determines that both the registration button switch 20 and the verification button switch 30 have not been operated in step 81 of the computer program started in step 80a according to the flowchart of FIG. Based on this, the determination process for the non-registration request and the non-verification request is repeated. In this state, when the operator presses his / her finger on the reading screen of the built-in image sensor, the fingerprint of the operator's finger is read by the image input device 50, and the image is read from the image input device 50. Fingerprint image information corresponding to the fingerprint image is generated as a fingerprint image signal.

When a registration operation signal is generated from the registration button switch 20 at such a stage, the microcomputer 60
However, in step 81, it is determined to be a registration request, and in step 82, the fingerprint image signal is received from the image input device 50,
At 83, the contents of the signal are binarized to form a binarized fingerprint image. In such a case, in each pixel of this binarized fingerprint image, each black (ie, “0”) pixel corresponds to each ridge of the one operator's fingerprint, while each white (ie, “1”) pixel. Pixels correspond to the valley lines of the fingerprint of the one operator. However, the valley line is
It corresponds to a valley between a pair of adjacent ridges of the fingerprint.

Then, in step 84, the microcomputer 60 smoothes the binarized fingerprint image by noise removal or the like to obtain a smoothed fingerprint image, and in step 85, thins the smoothed fingerprint image to perform minutiae extraction processing. At the same time, the pseudo minutiae removal process is performed. In such a case, the pseudo minutiae refers to an apparent minutiae that occurs due to ridge breaks, whiskers, adhesions, and the like. Thereafter, in step 86, the microcomputer 60 extracts the central image area excluding the peripheral image area from the image after the pseudo minutiae removal as described above (see FIG. 4) as the fingerprint area A, And each minutiae existing in this fingerprint area A (hereinafter, each minutiae M ₁ i (i = 0,1, ..., n))
Ask for. However, the number of pixels that specify the fingerprint area A is, for example, 512 × 512.

Thus, when the process in step 86 is performed, the microcomputer 60 classifies each minutia M ₁ i according to the distance from the center pixel of the smoothed fingerprint image in step 87a, and each minutia M ₂ i ( i = 0,1,
…, N) However, the pixel of the fingerprint area A has coordinates (x,
y), the central pixel is (x, y) = (255,255)
Specified by. Next, in step 87b, the microcomputer 60 checks the number of minutiae belonging to each range (Δx × Δy = 128 × 128) centered on each reception of the minutiae M ₂ i, and each minutiae M in descending order of the number. ₂ i is reclassified to obtain each minutia M ₃ i (i = 0, 1, ..., N).
In such a case, the result of the classification process in both steps 87a, 87b, each minutia M ₃ i is configured in the order a large number of minutiae that belong to each range around the minutiae. However, if the minutiae included are the same,
The one closer to the center of the fingerprint area A is prioritized.

When the minutiae classification process is completed as described above,
In step 88a, the microcomputer 60 assumes that each minutia M ₃ i corresponds to the first window area A ₁ , and if the image in the first window area A ₁ is blurred or crushed, for example, minutia is extracted. When it is determined that the number of pseudo minutiae is not large on the basis of the number of pseudo minutiaes obtained at times, a pixel region (128 × 128) centered on one minutia M ₃ k in each minutia M ₃ i is defined as the first window region A ₁ . Decide and register in the RAM. However, in the minutiae M ₃ k, the symbol k indicates i in the range of 0 ≦ k <n, and specifies the center of the first window region A ₁ .

Then, in step 88b, the microcomputer 60 assumes that each minutia M ₃ i in the range of k + 1 ≦ i ≦ n corresponds to the second window area A ₂ , and
When it is determined that the image blurring or blurring of the image of the wind area A ₂ is not large based on the number of pseudo minutiae obtained when extracting the minutiae, for example, the first wind area
The pixel area (for example, 32 × 64) centered on the single minutia M ₃ l in each minutia M ₃ i is determined as the second window area A ₂ so that the overlapping portion with A ₁ is reduced. Register with. However, in the minutiae M ₃ l, the symbol l indicates i in the range of k + 1 <l <n, and specifies the center of the second window region A ₂ (see FIG. 4). Further, the condition that the overlapping portion with the first window area A ₁ is reduced is the minutia M ₃ k in the first window area A ₁ .
This corresponds to a case where the pixel region 64 × 64 centered at the second window region A ₂ does not overlap. The pixel area 64 × 64 corresponds to the range of rough matching at the time of matching.

Thereafter, in step 88c, the microcomputer 60 assumes that each minutia M ₃ i in the range of l + 1 ≦ i ≦ n corresponds to the third window area A ₃ (see FIG. 4), and When there is not much blurring or crushing of the image in the window area A _3, the overlapping portion with the second window area A ₂ is reduced so that the pixel area centered on one minutia M ₃ m in each minutia M ₃ i is set to the first pixel area. 3 Determine the window area A ₃ and register it in the RAM. However, in the minutiae M ₃ m, the symbol m indicates i in the range of 1 + 1 <m <n, and specifies the center of the third window region A ₃ . In addition, the above-mentioned second
Condition that to reduce the overlapping portion of the window area A _2, when the pixel area 32 × 64 around the minutia M ₃ n in the second window area A ₂ overlap the third window region A ₃ is not satisfied Equivalent to. It should be noted that the remaining fingerprints of the respective operators are substantially the same as those described above in the first to third areas.
The window area of is determined and registered.

After that, when the one operator enters the special room, if the operator presses his / her finger on the reading screen of the built-in image sensor of the image input device 50 and operates the matching button switch 30 and the numeric keypad 40. , The image input device 50 generates the fingerprint image signal as described above, and the collation button switch
30 generates a verification operation signal, and the numeric keypad 40 generates a designated code signal. Then, the microcomputer 60
However, based on the collation operation signal, it is determined in step 81 that the collation request is made, and the computer program proceeds to the fingerprint collation routine 89. Then, this fingerprint matching routine 89
In accordance with the designated code signal, the microcomputer 60 reads out the first to third window areas corresponding to the contents of this signal from the RAM and compares them with the binarized image of the fingerprint image signal. If the verification results are in agreement, the door lock mechanism 10 is driven by the drive circuit 70 to unlock the door.

As described above, since the window area A ₁ of the first to third, as described above in the arithmetic process through steps 86~88c, A _2, A ₃ 5 is determined registered, the first window area A
₁ is the central portion of the fingerprint area where the fingerprint is less distorted and which contains more minutiae, and the second and third window areas A ₂ and A ₃ are
Next to the window area A ₁ , the image areas that are substantially different from each other and that are different from each other and that are different from the first window area A ₁ will be specified. Therefore, even if the binarized image has a low quality, each of the window regions A _{1 to} A ₃ can accurately specify the characteristic portion of the fingerprint. Also,
By using the window areas A _{1 to} A ₃ , it is possible to always perform highly accurate fingerprint matching even if the quality of the binarized image at the time of matching is low.

In the present embodiment, the number of pixels in the first window area A ₁ is 128 × 128. When the number of pixels is too large, the distortion increases and the matching recognition rate of the person decreases, and conversely This is because if it is too small, the false recognition rate with other people will increase. Also,
The reason why each of the window areas A _{1 to} A ₃ is rectangular is because it is convenient for hardware and software.

[Brief description of drawings]

FIG. 1 is a diagram corresponding to the description of the claims, FIG. 2 is a block diagram showing an embodiment of the present invention, FIG. 3 is a flow chart showing the operation of the microcomputer of FIG.
4 and FIG. 4 are exemplary binarization diagrams of fingerprint images. Explanation of symbols 20 …… Register button switch, 50 …… Image input device, 60…
… Microcomputer.

Front page continuation (72) Inventor Michicho Nagura, 1-1, Showa-cho, Kariya city, Aichi prefecture, Nihon Denso Co., Ltd. (72) Inventor, Koji Kawasaki, 1-1, Showa-cho, Kariya city, Aichi prefecture, Nihon Denso Co., Ltd. (56 ) Reference JP 59-151265 (JP, A) JP 60-15779 (JP, A) JP 63-78286 (JP, A) Actual development JP 62-187374 (JP, U)

Claims (1)

(57) [Claims] 1. An optical reading unit for optically reading a human fingerprint as a fingerprint image, a binarizing unit for binarizing the fingerprint image into a binarized image, and an image center of the binarized image. Extraction means for extracting each minutia in the area, initial classification means for classifying each minutia in the order of being closer to the center of the binarized image, and within the image central area centering each classification minutia by this initial classification means The number of minutiae belonging to each of the plurality of predetermined ranges, reclassifying means for reclassifying each of the classification minutiaes in descending order of the number of belonging minutiaes, and each of the reclassification minutiae by the reclassifying means, A predetermined range having the largest number of assigned minutiae in a predetermined range and a part of another predetermined range having the smaller number of assigned minutiae than the predetermined range are One predetermined range of reclassification minutiae and the other predetermined range of reclassification minutiae respectively as a center, so as to eliminate the overlap of a part within the one predetermined range and the part within the other predetermined range, A fingerprint registration device provided with determination registration means for determining the first and second window areas and for registering the first and second window areas.