JP2871157B2 - Fingerprint registration collation method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fingerprint registration and collation method, and more particularly, to a fingerprint identification and verification device such as an information security system. The present invention relates to a fingerprint registration / collation method for determining a use order at the time of collation.

In recent years, with the introduction of computers into a widespread society, concerned parties have been paying attention to system security. ID cards and passwords that have been used as a means of personal identification when entering a computer room or using a terminal have raised many questions in terms of security.

[0003] On the other hand, fingerprints are considered to be the most powerful means of personal identification because they have two major characteristics, "unique for everyone" and "invariant for life". R & D is ongoing.

[0004]

2. Description of the Related Art FIG. 10 is a block diagram of a conventional personal identification device, FIG. 11 is an explanatory diagram of a conventional "window" registration order, and FIG. 12 is a diagram showing examples of pseudo feature points.

In the figure, 1 is a fingerprint sensor, 2 is a binarization circuit,
3 is a binary memory, 4 is a registration unit, 5 is a feature information storage unit,
6 is a collating unit, 7 is a main displacement amount changing unit, 8 is a sub displacement amount changing unit, 9 is a memory address setting unit, 10 is a pixel value comparing unit,
Reference numeral 11 denotes a pass window evaluation unit, and reference numeral 12 denotes a feature extraction unit. Reference numeral 13 denotes an input image, P denotes a center of the input image, 14 denotes a valley portion of the fingerprint, and 15 denotes a ridge portion of the fingerprint.

Conventionally, a personal identification device based on the moving window method employs fingerprint sensors 1, 2 as shown in FIG.
It was composed of a binarization circuit 2, a binarization memory 3, a registration unit 4, a feature information storage unit 5, a collation unit 6, and the like.

The registration unit 4 includes a feature extraction unit 1
2 and the collation unit 6 includes a main displacement amount changing unit 7,
A sub-displacement amount changing unit 8, a memory address specifying unit 9, a pixel comparing unit 10, and a pass window number evaluating unit 11 are provided.

[0008] In the above device, when registering a fingerprint, the following is performed. First, a fingerprint image is input from the fingerprint sensor 1. This input image is binarized by the binarization circuit 2 and then stored in the binarization memory 3.

After that, the feature extracting unit 12 extracts a feature point of the fingerprint from the fingerprint image (binary image) in the binarization memory 3 and registers personal feature information as a dictionary in the feature information storage unit 5. .

At the time of this feature point extraction, a fingerprint image near a feature point such as an end point or a branch point of the fingerprint image is cut out as a “window”, and for each “window”, the position coordinates of the “window” and the “window” A binary image in the inside, a type of a feature point in the “window”, and the like are registered. At the time of registration as described above, “windows” are registered in the order shown in FIG.

That is, at the time of registration, the input image 13 is divided into several small areas, and the center P of the input image 13 is
In order from the small area close to
The method of searching for a feature point and registering it as a "window" was used.

The process at the time of fingerprint collation is as follows. First, a fingerprint image is input by the fingerprint sensor 1, a binarization process is performed by the binarization circuit 2, and the binarized fingerprint image is stored in the binarization memory 3.

Thereafter, the collating unit 6 collates the input image (binary fingerprint image) in the binarization memory 3 with the dictionary information in the feature information storage unit 5 to confirm the identity.

In this case, first, pattern matching between the "positioning window" and the input image is performed to correct the difference between the input position at the time of registration and the input position at the time of verification. Next, "Positioning window"
Is moved in parallel according to the amount of displacement of.

In these processes, the memory address output from the memory address specifying unit 9 is sent to the binarized memory 3 and the characteristic information storage unit 5 based on the main displacement amount set by the main displacement amount changing unit 7. It is done by doing.

By sending this address, the binary memory 3
And the window image information (dictionary information) from the feature information storage unit 5 are taken into the pixel value comparison unit 10 and the pixel values are compared.

However, in this comparison processing, the coincidence between the two pixels is not very good (because of the expansion and contraction of the finger, etc.). For this reason, the sub-displacement amount changing unit 8 causes each “collation window”
Are slightly moved independently of each other, and the pixel values are compared again.

"Window" which is information on the result of the pixel value comparison
Is passed to the number-of-accepted-windows evaluation section 11 to evaluate pass / fail. As a result of this evaluation, if more than a predetermined number of “windows” pass, an identity verification signal (a signal indicating that the identity has been confirmed) is output.

By the way, when registering a fingerprint, as described above, a method of sequentially searching for a feature point from a small area near the center of an input image and registering it as a "window" has been adopted.

For this reason, a “unique” “window”
Is not always selected preferentially, and pre-processing and post-processing for removing so-called pseudo feature points become complicated.

An example of the pseudo feature point is as shown in FIG. The example of FIG. 12A is an example in which a hole (in the rectangle in the figure) due to a sweat line is present on the ridge portion 15 of the fingerprint, so that it becomes a pseudo feature point.

That is, the part which is not the original feature point is a part (pseudo feature point) which cannot be distinguished from the original feature point by performing the thinning process and is extracted as the feature point.

In the example shown in FIG. 12B, since there is sweat on the valley line portion 14 of the fingerprint, a continuous pattern (within the dotted rectangle in the figure) between the two ridge portions 15 is formed. Are examples of pseudo feature points.

The processing for removing the pseudo feature points as described above has conventionally been performed by overlapping various mask processings.

[0025]

The above-mentioned conventional apparatus has the following problems. (1) At the time of fingerprint registration, feature points were searched in order from a small area near the center of the input image and registered as a "window". Therefore, "unique""windows" (real feature points) are not always selected with priority.

(2) For the reason (1), the selected “window” may include a pseudo feature point “window”. For this reason, it is necessary to remove the pseudo feature points, but this processing is complicated and the processing time is long.

(3) For the reasons (1) and (2), the reliability of the collation deteriorates. The present invention solves such a conventional problem, and when performing fingerprint registration, evaluates a “window” near an extracted feature point, and preferentially registers an original feature point that is not a pseudo feature point. The purpose is to do.

[0028]

FIG. 1 is a diagram showing the principle of the present invention, in which the same reference numerals as those in FIG. 10 denote the same parts. Reference numeral 20 denotes a use order setting unit, and reference numeral 21 denotes a matching frequency comparison unit.

The present invention has the following configuration in order to solve the above problems. (1) At least a feature information storage unit 5 that registers, as a dictionary, feature point information of a “window” cut out from the vicinity of a feature point of a fingerprint image, and a feature of a fingerprint registered in the feature information storage unit 5 And a collating unit 6 for collating with the input fingerprint feature
In the registration verification method of the personal identification device comprising: when registering a fingerprint, after registering a plurality of "windows" cut out from near the characteristic points of the input fingerprint image, for each of the plurality of "windows", Each of the registered “windows” is collated with a plurality of “windows” cut out from an input fingerprint image located in the vicinity, and the number of matches is counted (integrated). And the number of matches
The registration order, ordered in ascending order of
The use order of "windows" was set.

(2) When the count of the coincidence circuit exceeds a predetermined upper limit, the "window" is eliminated without performing the subsequent collation.

(3) When registering the plurality of "windows" ,
By performing similar feature point removal processing, the number of “windows” to be subsequently processed is reduced.

[0032]

(4) A weight coefficient is set and registered in the “window” according to the excessive number of times of matching, and at the time of fingerprint collation, identity verification is performed based on the magnitude of the total value of the weight coefficients.

[0034]

The operation of the present invention based on the above configuration will be described with reference to FIG. At the time of fingerprint registration, first, N (N is an arbitrary plural) “windows” are registered in the feature information storage unit 5 by the same method as the conventional method.

Thereafter, one "window" is selected from the registered "windows", and a plurality of "windows" cut out from the input image (in the binarization memory 3) located around the "window".
Are compared by the pixel value comparison unit 10, respectively.

As a result, if there is a matching "window", the number of times (the number of matches) is counted. Then, the number of matches is integrated for each registered “window” and stored in the memory within the use order setting unit 20.

When the above processing is completed for all "windows", the use order setting section 20 sorts the internally stored number of matches in ascending order.

Thereafter, the plurality of “windows” registered in the feature information storage unit 5 are ordered in ascending order of the number of matches.

In this way, the original feature points that are not the pseudo feature points are registered with priority. Since this ordering is used as a use order at the time of collation, stable identity verification can be performed in collation collation.

If the upper limit value is set as in the above configuration (2), unnecessary processing is not performed, and the processing can be sped up. Further, even when the number of “windows” to be processed is reduced as in the configuration (3), the processing can be sped up.

When the weight coefficient is set as in the above configuration (5), the identity verification process can be performed more stably than when the identity is simply verified by the number of passing windows.

[0042]

Embodiments of the present invention will be described below with reference to the drawings. (Explanation of First Embodiment) FIGS. 2 and 3 are views showing a first embodiment of the present invention, FIG.
FIG. 3 is a processing flowchart. 1 and 10 in the figures.
The same reference numerals denote the same components.

The personal identification device in the first embodiment is shown in FIG.
It is configured as shown in FIG. FIG. 2A is a block diagram of the personal identification device, and FIG. 2B is a data structure of personal characteristic information.

In the personal identification apparatus of this embodiment, a use order setting section 20 is provided in the registration section 4 of the conventional apparatus shown in FIG. 10, and a matching number counting section 21 is added to the collation section 6.

The personal characteristic information registered in the characteristic information storage unit 5 is composed of window position information (X, Y), use order information (window number), and window image information.

The coincidence counting section 21 counts (integrates) the number of coincidences for each "window" as a result of the pixel value comparison by the pixel value comparison section 10.

The use order setting unit 20 has a memory therein, and stores the number of matches for each “window” counted by the number-of-matches counting unit 21 in the memory. The order of the “windows” registered in the “windows” (“windows” registered in advance in the same manner as in the related art) is determined. Other configurations are the same as those of the conventional example shown in FIG.

Hereinafter, the processing of the first embodiment will be described with reference to the processing flowchart of FIG. Each processing number in FIG. 3 is shown in parentheses.

The fingerprint input image input by the fingerprint sensor 1 is converted into a binary image by the binarization circuit 2 (S1).
The data is stored in the binarization memory 3. Then, the feature extracting unit 12 thins the fingerprint image in the binarization memory 3 (S
2) is performed to extract feature points (S3).

Then, a "window" image is cut out from the vicinity of the feature point and registered in the feature information storage unit 5 as a dictionary. Note that these processes (S1 to S4) are the same as in the related art. By repeating the above process, N (N is an arbitrary integer other than 1) “windows” are registered.

Next, in the above-described processing, a process of evaluating the N “windows” registered in the characteristic information storage unit 5 and determining the use order (window number) at the time of comparison is performed. In this process, first, one “window” registered in the feature information storage unit 5 is taken out, and this “window” is collated with the input image stored in the binarization memory 3.

That is, the image information of the registered “window” and the image information of the “window” cut out from an arbitrary area of the input image are input to the pixel value comparing unit 10 to compare the pixel values. Do. In this case, for one registered “window”, “windows” cut out from the input image are a plurality of “windows” set in a certain search range (a plurality of “windows” at different locations). A plurality of "windows" are sequentially compared.

As described above, the registered “window”
Must be selected, and one “window” needs to be cut out from the input image in the binarization memory 3 located around the “window”.

Therefore, the "window" to be cut out from the input image
Is set (referred to as an auxiliary displacement amount) (S
5). The setting of the displacement amount is performed by the sub displacement amount changing unit 8.

Thereafter, if the displacement is within a predetermined search range (S6), one "window" registered in the feature information storage unit 5 by the pixel value comparison unit 10 and the binarized memory For the “window” cut out from the input image in No. 3, pixel comparison is performed by comparing the pixel values in each “window” (S7).

By the above-mentioned pixel collation, the registered “window”
(Temporarily, a temporary dictionary) and the degree to which "window" cut out from the input image matches (S8). This process is performed, for example, by detecting the number of times that the degree of inconsistency falls below a predetermined threshold value for the pixel values in both “windows” to be compared.

If the number of times is equal to or less than a predetermined value, the two "windows" being compared are determined to be non-matching "windows". Is determined to be. It should be noted that the determination as to whether or not they match is made irrespective of the determination at the time of comparison (not necessarily the same).

If the two “windows” match as a result of the above pixel value comparison, the number of matches is counted by the matching number counting section 2.
It counts with 1 (S9). If they do not match, the counting is not performed.

Next, the selected registered “window” is left as it is, and the “window” cut out from the binarization memory 3 is changed to another “window” (by changing the sub-displacement amount). The comparison is performed, and if they match, the number of matches is integrated (S5 to S5).
S9).

In this way, one registered “window”
On the other hand, collation is performed while sequentially moving “windows” of the input image located in the vicinity thereof, and the number of times (the number of windows) when both “windows” match each other is accumulated.

When this process is completed within the predetermined search range (S6), another one of the registered "windows" is selected again, and this "window" is compared with the input image. I do. Also in this case, the number of matches is integrated in the same manner as described above.

At this time, an integrated value of the number of matches is calculated for each registered “window”, and this value is stored in the memory in the use order setting unit 20.

When the integrated value of the number of matches is obtained for all registered "windows" (S10), the use order setting unit 20 orders the "windows" (S11) and registers them (S11).
12).

In this processing, the number of matches is compared for each “window”, the numbers are arranged in ascending order of the number of matches, and the use order information (window number) is written in the feature information storage unit 5 for processing.

As a result, when the number of matches is small, there are few similar patterns around the “window”, and it is highly likely that the registered “window” is not a pseudo feature point (original feature point). .

Therefore, if the “windows” are registered and arranged in ascending order of the number of matches, and this order is used as the order of use at the time of fingerprint collation, it is possible to perform collation using true feature points from which pseudo feature points have been removed. .

(Explanation of the Second Embodiment) FIG. 4 is a block diagram of the personal identification device according to the second embodiment. FIG. 4 (A) is a block diagram of the personal identification device, and FIG. 4 (B) is personal characteristic information. FIG. 3 is a diagram showing a data structure of FIG. FIG. 5 is a processing flowchart of the second embodiment. In the figure, the same reference numerals as those in FIG. Reference numeral 22 denotes an upper limit setting unit for the number of matches.

The personal identification device of the second embodiment is obtained by adding an upper limit setting unit 22 for the number of matches to the device shown in FIG. 2, and the other configuration is the same as that of FIG.

The upper limit setting unit 22 for the number of coincidences sets an upper limit to the number of coincidences counted by the number of coincidence counting unit 21. By setting this upper limit, "windows" which are apparently not the original feature points are eliminated at an early stage.

In this embodiment, the characteristic information storage unit 5
The personal characteristic information registered in the window position information (X,
Y), use order information (window number), window image information (binary)
It consisted of. The use order information (window number) is data used as a use order at the time of collation.

Next, the processing of the second embodiment will be described with reference to the processing flowchart of FIG. Each processing number in FIG. 5 is shown in parentheses. Each process in FIG. 5 (S20 to S32)
Among them, the processing other than the processing of S29 is the same as the processing of the first embodiment shown in FIG.

In this embodiment, when the number of matches is counted (integrated) in the process of S28, if the number of matches exceeds the upper limit value set by the upper limit setting unit 22 of the number of matches (S28).
29) Then, the collation for the “window” is terminated (excluded).

That is, if the number of matches is large, the “window”
There is a high possibility that it is not an original feature point but a pseudo feature point. Therefore, when the number of matches exceeds the upper limit, the collation for that “window” is terminated, and the process proceeds to the collation of the next “window”, so that useless processing can be reduced and the processing time as a whole becomes shorter.

(Explanation of Third Embodiment) FIG. 6 is a block diagram of a personal identification device according to a third embodiment. FIG. 6A is a block diagram of the personal identification device, and FIG. 6B is personal characteristic information. FIG. 3 is a diagram showing a data structure of FIG. FIG. 7 is a processing flowchart of the third embodiment. In the figure, the same reference numerals as those in FIG. Reference numeral 23 denotes a pseudo feature elimination unit.

The third embodiment is an example in which a pseudo feature elimination unit 23 is added to the registration unit 4 of the personal identification device shown in FIG. 2, and the other configuration is the same as that of the device of FIG.

In the third embodiment, the personal characteristic information is
It consisted of window position information, use order information (window number), and window image information (binary).

When registering a plurality of “windows” by the conventional method, the pseudo information feature elimination unit 23 performs a simple pseudo feature point exclusion process, thereby performing the subsequent processing of the “window”. The number should be reduced.

Hereinafter, the processing of the third embodiment will be described with reference to the processing flowchart of FIG. Each processing number in FIG. 7 is shown in parentheses.

The processing of the third embodiment is the same as the processing of the first embodiment shown in FIG. 3 except for the processing of S43 during the processing of S40 to S52.

As shown in the figure, after performing the feature extraction (S42), simple removal of pseudo feature points (S43) is performed. This process is performed by the pseudo feature elimination unit 23. In this case, the removal of the pseudo feature point is performed by a simple process such as “crack removal” (described in the conventional example) only.

By this processing, pseudo feature points are removed to some extent, the number of “windows” registered in the feature information storage unit 5 is reduced, and the subsequent processing time is shortened.

(Explanation of the Fourth Embodiment) FIG. 8 is a block diagram of the personal identification device of the fourth embodiment. FIG. 8A is a block diagram of the personal identification device, and FIG. 8B is personal characteristic information. FIG. 3 is a diagram showing a data structure of FIG. FIG. 9 is a processing flowchart. In the figure, the same reference numerals as those in FIG.

The identity verification device of the fourth embodiment is the same as the device shown in FIG. In addition, data of personal characteristic information
It is composed of window position information (X, Y), use order information (window number), weight coefficient, and window image information (binary).

Hereinafter, the processing of the fourth embodiment will be described with reference to the processing flowchart of FIG. Each processing number in FIG. 9 is shown in parentheses.

The processes in S60 to S72 in FIG. 9 are the same as those in the first embodiment shown in FIG. 3, except for the process in S71.

In this embodiment, after the "window" ordering process (S70), a weighting factor is set (S71) and registered (S72).

In this case, the use order setting section 20 stores the number of matches (integrated value) for each “window” in the internal memory, and performs the ordering of the “windows” (in order of decreasing number of matches). After that, the weight coefficient K ₁ (see FIG. 8) is set as personal characteristic information in the characteristic information storage unit 5.

As a method of setting the weight coefficient, for example, a comparison table between the number of matches and the weight coefficient is stored in a memory in the use order setting unit 20 in advance, and each “ For each “window”, a weighting factor is set according to the number of matches.

The weight coefficient set as described above is used at the time of fingerprint collation. At the time of fingerprint collation, the number of acceptable windows is evaluated in the number of acceptable windows evaluation unit 11, and in this case,
Rather than outputting an identity confirmation signal (determination signal) only by the number of pass windows, processing is performed in consideration of the above-mentioned weighting factors. For example, identification is performed based on the magnitude of the total value of the weight coefficients.

(Other Embodiments) Although the embodiments have been described above, the present invention can be implemented as follows. (1) In the first embodiment, the process of S9 in FIG. 3 may count the number of positions where the degree of mismatch is lower than a predetermined threshold. In this case, in the “window” ordering process in S11, the ordering is performed in ascending order of the number of times that the degree of inconsistency falls below the threshold value. By performing such processing, the processing can be speeded up as compared with the case where the number of matches is counted.

(2) In the third embodiment, S4 in FIG.
In the case of performing the third process, the removal of the pseudo feature point may be performed by only one type of process, or may be performed by two or three types of processes. In any case, all the pseudo feature points need not be removed, but those which can be easily processed may be selected.

[0092]

As described above, the present invention has the following effects. (1) Since the original feature points that are not the pseudo feature points are registered with priority, the reliability of fingerprint collation is improved.

(2) Since it is not necessary to perform the complicated removal processing of the pseudo feature points as in the related art, the processing at the time of fingerprint registration is simple and the speed can be increased.

(3) If processing is performed with the upper limit of the number of matches set, processing at the time of registration is further accelerated.

(4) If a weighting factor is set according to an excessive number of matchings, identity verification at the time of collation can be performed more stably.

[Brief description of the drawings]

FIG. 1 is a principle diagram of the present invention.

FIG. 2 is a configuration diagram of the personal identification device according to the first embodiment of the present invention.

FIG. 3 is a processing flowchart of the first embodiment.

FIG. 4 is a configuration diagram of a personal identification device according to a second embodiment.

FIG. 5 is a processing flowchart of a second embodiment.

FIG. 6 is a configuration diagram of a personal identification device according to a third embodiment.

FIG. 7 is a processing flowchart of a third embodiment.

FIG. 8 is a configuration diagram of an identity verification device according to a fourth embodiment.

FIG. 9 is a processing flowchart of a fourth embodiment.

FIG. 10 is a configuration diagram of a conventional personal identification device.

FIG. 11 is an explanatory view of a conventional “window” registration order.

FIG. 12 is an example of a pseudo feature point.

[Explanation of symbols]

 Reference Signs List 3 Binarized memory 4 Registration unit 5 Feature information storage unit 6 Collation unit 10 Pixel value comparison unit 12 Feature extraction unit 20 Usage order setting unit 21 Matching number counting unit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G06T 7/00

Claims (4)

(57) [Claims] 1. A feature information storage unit (5) for registering at least feature information of a “window” cut out from the vicinity of a feature point of a fingerprint image as a dictionary, and a feature information storage unit (5). In the fingerprint registration / collation method of the personal identification device having a collation unit (6) for collating the characteristics of a given fingerprint with the characteristics of the input fingerprint, when the fingerprint is registered, it is cut out from the vicinity of the characteristic point of the input fingerprint image. After registering a plurality of “windows”, each of the plurality of “windows” is checked against a plurality of “windows” cut out from an input fingerprint image located in the vicinity thereof, and the number of matches is counted (integrated). ) and, in ascending order of the number of matches (integrated value), the ordering of the "window" registered above, the registration order which is ordered in ascending order of the number of coincidences, the finger
A fingerprint registration / collation method characterized by using the order of use of “windows” during pattern collation . 2. The fingerprint registration and collation according to claim 1, wherein when counting the number of matches, when a predetermined upper limit value is exceeded, the "window" is excluded without performing subsequent collation. Method. 3. The method according to claim 1, wherein, when registering the plurality of “windows”, the number of “windows” to be subsequently processed is reduced by performing pseudo feature point removal processing. Fingerprint registration collation method described. 4. A method according to claim 1, wherein a weighting factor is set and registered in the "window" according to the excessive number of matchings, and at the time of fingerprint collation, identity verification is performed based on the magnitude of the sum of the weighting factors. The fingerprint registration / collation method according to claim 1.