JPH05108806A - Picture characteristic extracting method and device - Google Patents

Abstract

PURPOSE:To extract even the feature point of a part where the flow of a rising line pattern is rapidly changed by combining corresponding minutiae so that the edge point of the corresponding ridge line can be a pair with the branch point of the corresponding valley line, and extracting the feature point by the pair groups of the minutiae. CONSTITUTION:The ridge minutiae corresponds to the valley by one to one. That is, the edge point of the ridge appears corresponding to the branch point of the valley, and the branch point of the ridge appears corresponding to the edge point of the valley. Then, a corresponding relation between the edge point of the ridge and the branch point of the valley is checked (231) by using the obtained minutiae list, so that the correspondence can be related, and the specific independent minutiae can be extracted. In the same way, the corresponding relation between the edge point of the valley and the branch point of the ridge is checked (234). Then, for example, data obtained by averaging the minutiae position or the direction are used as the characteristic amounts of the minutiae based on the corresponding minutiae list obtained by a corresponding minutiae extracting means 240.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image feature extraction method and apparatus used for collation or classification of skin patterns such as fingerprints.

[0002]

2. Description of the Related Art A fingerprint collation device used for entrance control, criminal investigation and the like collates and classifies fingerprint patterns. At this time, it is important how to extract characteristics of fingerprint patterns. When collating fingerprints, minutiae consisting of fingerprint ridge end points and branch points are used for this feature extraction. Further, when classifying fingerprints, singular points such as the center of a fingerprint pattern and a delta are used. Feature extraction for skin ridges such as finger prints and palm prints can be handled in the same way as fingerprints. Here, the method of extracting the minutiae of the fingerprint,
Two points of a method of extracting a singular point which is a portion where the flow of a pattern changes abruptly will be described.

Conventionally, when a fingerprint minutia is extracted, a fingerprint collation device first binarizes an input fingerprint pattern image and further thins it. Next, the end points and branch points of the thin line image thus obtained are extracted as minutiae of the fingerprint. A method of extracting this minutiae from a striped pattern such as a skin pattern is an image in which the ridge side of the fingerprint is thinned (ridge thin line image) and an image in which valleys appearing between the ridges are thinned (valley). Thin line image) (Japanese Patent Application Laid-Open No. 59-144982). In this method, first, a mask is set according to the direction of the ridge and the end point of the valley, and in the mask, there is a valley branch point for a ridge, and a ridge branch point for a valley. If so, the end points of the ridge are selected, and if they are valleys, those without the end points of the valley are selected as feature points. Although the ridge end points and branch points should be selected as feature points, the ridge branch points are regarded as corresponding to the valley end points, and the ridge and valley end points are selected as feature points. The ridge direction is obtained by tracing the ridge from the end points of the ridge and valley. That is, this method uses the skin pattern-like property that the branch points of the ridge correspond to the end points of the valley, and the branch points of the valley correspond to the end points of the ridge, as shown in the skin pattern of FIG. 7. There is. However, this property does not always hold for the center of a fingerprint or the minutiae of a delta where the flow of ridges changes rapidly. That is, as shown in FIG. 8, there is a minutia in the ridge image, but the minutia corresponding to the minutia does not appear in the valley image, and as shown in FIG. 9, the minutia exists in the valley image, but there is a minutia in the ridge image. There is a case where the minutiae does not appear. Therefore, when this method is used, the feature points in the portion where the ridge flow changes abruptly may be missed.
The method of extracting minutiae from a skeletal line image has been described above. However, extracting minutiae does not necessarily require fine lines, and there is also a method of extracting minutiae from grayscale images (Koichiro Morita et al. IEICE, AL85-56, "Fingerprint collation terminal for personal identification").

As a method of detecting the center of a fingerprint pattern or a delta, a method using a ridge direction pattern is generally used. One of the methods is to detect it by examining the direction distribution (IEICE Transactions, '8.
2, vol. J65-D. No. 10. pp1286-
1293, N. Nakamura et al., "Classification of Fingerprint Images by Directional Distribution Pattern"). In this method, first, the fingerprint image is divided into small areas, and the ridge direction in each small area is obtained to extract the ridge direction pattern of the entire fingerprint. Next, a circular window is passed through this ridge direction pattern and the direction distribution in this region is obtained to detect singular points such as the center and delta. However, when a singular point such as a center or a delta is detected using a ridge direction pattern, it is not possible to indicate on which ridge the singular point exists because the ridge image is not referenced.

[0005]

According to the above-mentioned conventional method, when the minutiae on the ridge line and the minutiae on the valley line are compared, even if the image is ideally beautiful, the center of the fingerprint or the ridge of the skin in the delta shape is compared. In the portion where the flow of changes drastically, the correspondence between the end points and the bifurcation points does not always hold in the minutia extracted from the ridge line and the minutia extracted from the valley line. That is, even a minutia that can be extracted with a ridge line, a minutia corresponding to the minutia of the ridge line is not extracted with a valley line, or a minutia extracted with a valley line is not extracted with a ridge line. Further, it is impossible to detect a singular point in the center of the fingerprint or a delta-shaped portion. Furthermore, even in the method of extracting a minutia using a ridge image and a valley image, the ridge direction and position are extracted from one minutia extracted, and the corresponding minutia direction and position are ignored. Will be done.

An object of the present invention is to more stably extract the features of minutia by obtaining the correspondence of minutiae, and at the same time, abruptly generate a ridge-like pattern which cannot be extracted by the conventional method. It is also to extract the feature points in the changing part. Further, the singular point in the center of the fingerprint or the delta-shaped portion is extracted so that the minutiae information can be used for fingerprint, palmprint classification, fingerprint collation device or personal identification device.

[0007]

According to a first aspect of the present invention, there is provided an image feature extraction method for extracting features from a skin pattern image, which comprises a minutiae group of ridge lines corresponding to ridges of the skin pattern. Extract the valley minutiae group corresponding to the portion of the valley between the ridges, of the ridge line minutiae group and the valley line minutiae group, of the corresponding ridge line end points and valley line branch points. It is characterized in that corresponding minutiae that form a pair or a pair of a branch point of a corresponding ridge line and a pair of end points of a valley line are combined, and feature points of an image are extracted by a group of minutiae pairs.

The image feature extracting method of the second invention is an image feature extracting method for extracting features from a skin pattern image.
A group of minutiae on the ridge line corresponding to the ridge of the skin pattern,
A valley line minutiae group corresponding to a valley portion between the ridge lines is extracted, and the corresponding ridge line end point and valley line branch point among the ridge line minutiae group and the valley line minutiae group. , Or a pair of corresponding minutiaes that form a pair of branch points of the corresponding ridge line and end points of the valley line, and an isolated peculiar minutiae without corresponding minutiae is extracted as a feature point of the image. To do.

The image feature extraction method of the third invention is an image feature extraction method for extracting features from a skin pattern image.
A group of minutiae on the ridge line corresponding to the ridge of the skin pattern,
A valley line minutiae group corresponding to a valley portion between the ridge lines is extracted, and the corresponding ridge line end point and valley line branch point among the ridge line minutiae group and the valley line minutiae group. , Or a pair of corresponding minutiae that form a pair of branch points of the corresponding ridge line and end points of the valley line, feature points of the image are extracted by the group of minutiae, and isolated singularities without the corresponding minutiae. Image feature extraction method for extracting various minutiae as feature points of an image.

The image feature extracting apparatus of the fourth invention is a ridge minutia extracting means for extracting a minutia group of ridge lines corresponding to ridges of a skin pattern image, and a valley corresponding to a valley portion between the ridges. A valley minutia extracting means for extracting a minutia group of lines, and a pair of a ridge line end point and a valley line branch point of the extracted ridge line minutia group and valley line minutia group, or a corresponding ridge line And corresponding minutia extracting means for extracting corresponding minutiae that form a pair of the branch point and the end point of the valley line.

An image feature extracting apparatus of a fifth invention is a ridge minutia extracting means for extracting a minutia group of ridge lines corresponding to ridges of a skin pattern image, and a valley corresponding to a valley portion between the ridges. A valley minutia extracting means for extracting a minutiae group of lines, a pair of a ridge line minutiae group and a valley line minutiae group, a pair of corresponding ridge line end points and valley line branch points, or a corresponding ridge line The present invention is characterized by comprising an isolated minutia extracting means for combining corresponding minutiaes that form a pair of a branch point and an end point of a valley line and extracting an isolated minutiae without a corresponding minutiae.

[0012]

[Function] From the skin pattern, to the part where the flow of ridges changes rapidly (the center or delta in the case of fingerprints) can be detected with the ridge line but not with the valley line, or with the minutia or valley line There may be minutiae that are extracted but not on the ridge line. The present invention can extract the minutiae as feature points. Further, with respect to the minutiae that can be extracted, the minutiae of the corresponding ridge and valley can be usually extracted, so that the direction and position of the minutiae can be extracted more accurately.

This will be described below.

The minutia appearing in a portion where the skin pattern flows gently is, for example, if it is an end point in a ridge image as shown in FIG. 7, it appears as a branch point in a valley image. In this way, since the branch point of the valley appears near the end point appearing in the ridge image, this minutia combination is extracted as a feature point. On the contrary, the minutia that appears as an end point in the valley image appears as a branch point in the ridge image.
In this case as well, minutiae can be similarly extracted. But,
This does not apply to the part where the flow of the ridge changes rapidly. First, as an example thereof, a portion in which the flow of skin ridges sharply changes in the center of a fingerprint will be described with reference to the images of FIGS. 8 and 9. In FIG. 8, first, nothing is detected as the ridge minutiae. Detect endpoints as valley minutiae. Since there is no minutia near this end point, it is determined that there is no corresponding minutia,
This point is extracted as a singular point. On the contrary, in FIG. 9, even if there is an end point at the ridge, there is no branch point at the valley.

The part where the flow of ridges changes like a delta of a fingerprint will be described with reference to the image of FIG. 10 as an example.
In this image, first, as a ridge minutia,
No minutiae are detected. In the valley image, the branch point can be detected. Since the minutiae of the corresponding ridge are not detected in the vicinity of this branch point, this branch point is extracted as a singular point.

As described above, according to the present invention, feature points that cannot be detected by examining only the ridge line image or only the valley line image can be extracted, and the flow of ridges can be sharp. It is also possible to extract a singular point in a portion that changes to. This makes it possible to extract singular points in the core and delta that are conventionally used for classification of fingerprint images.

Since the correspondence between the ridge minutiae and the valley minutiae can be obtained, the minutiae information can be obtained by averaging the information such as the position and direction extracted from the corresponding end points and branch points. It can be obtained stably. In particular, the direction information can be obtained more accurately in a portion where the flow of the ridge changes relatively rapidly.

[0018]

FIG. 1 is a block diagram showing an embodiment of an image feature extracting apparatus of the present invention.

This image feature extraction device stores an image storage means 11 for recording an image of a skin pattern such as a fingerprint input from a CCD camera or the like, and a ridge image corresponding to a ridge portion of the stored skin pattern image. Ridge minutia extraction means 12 for extracting minutiae which are the end points or branch points of the ridge line.
And a valley minutia extracting means 13 for extracting a minutiae which is an end point or a branch point of the valley line of the valley image corresponding to the valley portion between the ridges, and the ridge between the extracted ridge minutiae and the valley minutiae. Corresponding minutia extraction means 14 for extracting minutiae having a corresponding relationship such as a valley branch point for an end point and a valley end point for a ridge branch point.
And minutia feature extraction means 15 for extracting minutia features such as the direction of the minutia from the extracted minutia.

A case will be described in which the image feature extracting device having the above-described configuration extracts feature points of a skin pattern. Figure 1
In order to explain in more detail, the case where processing is performed by the processing flow as shown in FIG. 2 will be described. The image input by the CCD camera or the like is stored in the image storage unit 200. The image storage means is realized by a memory.

The ridge minutia extraction means 210 for extracting the ridge minutiae from the image stored in the image storage means 200 binarizes and cores this image, and then scans the image with a minutiae detection mask as shown in FIG. To detect minutiae. The minutiae detected by the minutiae detection mask shown in FIG. 4A are end points, and the minutiae detected by the mask shown in FIG. 4B is a branch point. In this way, the ridge minutia extraction means 210 extracts the end points of the ridge image and the branch points of the ridge image, respectively, and creates a list thereof (steps 211 and 212).

The valley minutia extracting means 220 for extracting the valley minutia is the ridge minutia extracting means 2.
By inputting an image obtained by reversing the shade of the ridge image as the input image of 10, the same processing as the ridge image is performed. In addition, ridge and valley minutiae extraction means 210, 220
Does not necessarily require skeletonization, and uses a raster matching method that directly extracts minutiae from grayscale images (Koichiro Morita et al., IEICE Study Group Material AL85-
56, "Fingerprint collation terminal for personal identification").

Next, the corresponding minutia extracting means 230 for associating the minutia groups of the ridge and the valley thus extracted will be described. Ridge minutiae and trough minutiae usually have the following one-to-one correspondence.
The end points of the ridge appear in correspondence with the bifurcation points of the valley, and the bifurcation points of the ridge appear in correspondence with the end points of the valley. Correspondence is obtained using this property. Steps 231, 232, 23
In step 3, the correspondence between the ridge end points and the valley bifurcation points is examined using the minutialists obtained in steps 211 and 222. By this processing, the correspondence between the end points of the ridge and the branch points of the valleys is associated with each other, and the unique minutiae isolated in each processing step are extracted. Similarly, in steps 234, 235, and 236, the correspondence between the valley end points and the ridge bifurcation points is also checked. Based on the corresponding minutialist obtained by the corresponding minutia extracting means, in step 240, for example, by averaging minutiae positions and minutiae directions as minutia feature amounts, conventional minutiae network feature etc. feature extraction methods (electronic IEICE Transactions D-II, vol. J72-D-II, No.
5, pp. 724-732 (1989), Hiroshi Asai and others,
"Automatic fingerprint matching by minutia network features-
Features are extracted by the feature extraction process- ").

In the following, steps 231, 232, 233
Will be described in detail. Step 231 is realized by an algorithm as shown in FIG.

First, in step 301, a valley branch point appearing within a distance L near the ridge end point i registered in the ridge end point list obtained by the process 211 is selected from the valley branch point list. look for. This calculates the distance from the coordinates of the end points and the coordinates of the branch points, and judges whether or not it is the neighborhood by the distance.However, mask processing is performed centering on the end points, and the existence of the branch points in the mask causes It may be determined whether or not there is a branch point. When there is no valley branch point near the ridge end point i, it is determined that the ridge end point i does not have a corresponding valley branch point. Then, the end point i is set as an isolated point and registered in the isolated point list in step 311. If there is a trough branch point near the ridge end point i, the trough branch point is registered in the neighbor branch point list of the end point i in step 302. Next, regarding the branch point in the registered list of neighboring branch points, in step 303, the end point i
Check whether there is a ridge line or a valley line between and the branch point. If a ridge line or a valley line exists between the end point i and the branch point, it is determined that this branch point is not the valley branch point having a corresponding relationship with the end point i, and the branch on the remaining nearby branch point list is determined. Examine the points. If the ridge line or the valley line does not exist between the end point i and the branch point, in step 304, this valley branch point is set as the candidate of the valley branch point corresponding to the ridge end point, and the corresponding branch point candidate is selected. Register in the list. Step 305
Then, after checking for the existence of a ridge line and a valley line for all the neighboring branch points on the nearby branch point list, in step 306, the number of corresponding branch point candidates on the corresponding branch point candidate list is checked. When there is a ridge line or a valley line between all the branch points on the nearby branch point list and the end point i, and there is no corresponding branch point candidate, the end point i of the ridge is determined to be an isolated point. When there are a plurality of corresponding branch point candidates, this corresponding branch point list is stored as a corresponding branch point list for the ridge end point i. If only one corresponding branch point exists, it is determined that the end point i and the only trough branch point j on the corresponding branch point list are the branch points having a corresponding relationship with the end point of the ridge, and in step 314, the corresponding minor point is selected. register. Since it is not necessary to check the correspondence between the registered ridge end point i and the valley branch point j thereafter, it is excluded from the valley branch point list when searching for a nearby branch point. If this branch point j exists on the corresponding branch point candidate list of the end point m of another ridge, it is also excluded from this list. As a result, if the only candidate in the corresponding branch point candidate list of the ridge end point m is the valley branch point n, it is determined that the ridge end point m and the valley branch point n have a corresponding relationship. .. After that, similarly returning to step 313, the end point m of the ridge and the branch point n of the valley are registered in the corresponding minutialist and the list is updated. After registering the corresponding minutiae and updating the list, the process returns to step 309. If the process has not been performed for all the ridge end points, the process returns to step 302 to repeat the process. When the processing is completed for the end points of all ridges, the next processing step is performed. With the above, step 231 of FIG. 2 is completed.

After step 231 is completed, the routine goes to step 232. This is because even after the processing for all the end points is completed in step 231, a plurality of corresponding branch point candidates for the two end points remain forever, for example, in the case of a broken minutia as shown in FIG. Will end up. In order to deal with such a case, when there are a plurality of processing corresponding branch point candidates, among the branch points on the corresponding branch point candidate list, the corresponding branch point candidate closest to the end point of the ridge is determined to be the corresponding branch point, Register with the corresponding minutialist. After registration, the list is updated in the same manner as in step 313 and subsequent steps in FIG. In this way, the endpoints of all ridges are registered in the corresponding minutialist or isolated point list. The remaining valley branch points that are not registered in the corresponding minutialist are all determined to be isolated points in step 233.

FIG. 6 is a block diagram showing another embodiment of the image feature extraction apparatus of the present invention.

This image feature extracting apparatus includes an image storing means 61 for storing an image of a skin pattern such as a fingerprint input from a CCD camera or the like, and a ridge of a ridge image corresponding to a ridge portion of the stored skin pattern image. Ridge minutia extraction means 62 for extracting minutiae which are end points and branch points of the line,
A valley minutia extracting means 63 for extracting a minutia which is an end point or a branch point of a valley line of a valley image corresponding to a valley portion between ridges, and a valley if it is an end point of the ridge between the extracted ridge minutiae and the valley minutiae. A branch point and an isolated minutia extracting means 64 for extracting minutiae that do not correspond to each other, such as a valley end point for a ridge branch point, and an isolated minutia utilizing means for utilizing the isolated point extracted by the isolated minutia extracting means are provided. Prepare

The means 61, 62 and 63 in FIG. 6 are the same as those in FIG.
Nos. 11, 12, and 13 are the same. The isolated minutia extraction means 65 outputs the isolated points obtained by the intermediate processing of the corresponding minutia extraction means 14 of FIG. The isolated minutia utilization means 65 is means for utilizing the singular minutiae on the isolated point list output by the isolated minutia extraction means 64. For example, the singular minutia is CR.
It is realized by displaying on T. The extracted singular minutiae are singular points related to the fingerprint center and deltas that are used as references when classifying fingerprints. By using the isolated minutia utilization means 65 to superimpose and display this peculiar minutia on, for example, the original image or the core line image, it becomes possible to easily perform manual classification of fingerprints.

[0030]

According to the present invention, in addition to minutiae extracted from only a ridge image or only a valley image, minutiae that could not be obtained independently can be extracted. As a result, many feature points can be extracted, and fingerprint matching accuracy can be improved.

Further, by obtaining the correspondence between minutiae between ridges and valleys, it becomes possible to more stably obtain minutiae information such as the position and direction of the corresponding minutiae. The accuracy of fingerprint matching can be improved.

Further, according to the present invention, the singular points appearing in the abruptly changing portion of the skin pattern can be extracted independently of the normal minutiae. As a result, when the fingerprints are classified, it is possible to extract the singular points appearing in the important fingerprint center and the delta portion, and it is possible to obtain the effect of facilitating the classification of the fingerprints. It goes without saying that the classification of fingerprints is performed at the time of registration and verification in the fingerprint verification device used in criminal scanning.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an image feature extraction device of the present invention.

FIG. 2 is a diagram for explaining FIG. 1;

FIG. 3 is a diagram for explaining corresponding minutia extraction means in FIG.

FIG. 4 is a diagram for explaining minutiae detection in FIG. 1.

5 is a diagram for explaining the corresponding minutiae in FIG. 1. FIG.

FIG. 6 is a block diagram showing an embodiment of an image feature extraction device of the present invention.

FIG. 7 is a diagram for explaining the operation of the present invention.

FIG. 8 is a diagram for explaining the operation of the present invention.

FIG. 9 is a diagram for explaining the operation of the present invention.

FIG. 10 is a diagram for explaining the operation of the present invention.

[Explanation of symbols]

 11 image storage means 12 ridge minutia extraction means 13 valley minutia extraction means 14 corresponding minutia extraction means 15 minutia feature extraction means 61 image storage means 62 ridge minutia extraction means 63 valley minutia extraction means 64 isolated minutia extraction means 65 display means

Claims (5)

[Claims] 1. An image feature extraction method for extracting features from a skin pattern image, wherein a group of minutiae of ridge lines corresponding to the ridges of the skin pattern and a minutiae of valley lines corresponding to a valley portion between the ridges. A group of the ridge line minutiae and the valley line minutiae group, a pair of corresponding ridge line end points and valley line branch points, or a corresponding ridge line branch point and valley line end points. An image feature extraction method, characterized in that corresponding minutiae that form a pair of the minutiae are combined and feature points of the image are extracted by the group of minutiae pairs. 2. An image feature extraction method for extracting features from a skin pattern image, wherein a group of ridge line minutiae corresponding to the skin pattern ridges and a valley line minutiae corresponding to a valley portion between the ridge lines. A group of the ridge line minutiae and the valley line minutiae group, a pair of corresponding ridge line end points and valley line branch points, or a corresponding ridge line branch point and valley line end points. An image feature extraction method characterized by combining corresponding minutiae that form a pair of and extracting an isolated peculiar minutiae without corresponding minutiae as feature points of the image. 3. An image feature extraction method for extracting features from a skin pattern image, wherein a group of ridge line minutiae corresponding to the skin pattern ridges and a valley line minutiae corresponding to a valley portion between the ridge lines. A group of the ridge line minutiae and the valley line minutiae group, a pair of corresponding ridge line end points and valley line branch points, or a corresponding ridge line branch point and valley line end points. An image feature extraction method of combining corresponding minutiae that form a pair of the minutiae, extracting feature points of the image by the group of minutiae, and extracting an isolated peculiar minutiae without the corresponding minutiae as the feature point of the image. 4. A ridge minutia extracting means for extracting a minutia group of ridge lines corresponding to ridges of a skin pattern image, and a valley minutiae for extracting a minutia group of valley lines corresponding to a valley portion between the ridge lines. Of the extracted ridge line minutiae group and valley line minutiae group, the corresponding ridge line end point and valley line branch point pair, or the corresponding ridge line branch point and valley line end point pair An image feature extracting apparatus, comprising: corresponding minutia extracting means for extracting corresponding minutiae that form a pair. 5. A ridge minutia extracting means for extracting a minutia group of ridge lines corresponding to ridges of a skin pattern image, and a valley minutiae for extracting a minutia group of valley lines corresponding to a valley portion between the ridges. Extraction means, of the ridge line minutiae group and the valley line minutiae group, a pair of corresponding ridge line end points and valley line branch points, or a pair of corresponding ridge line branch points and valley line end points An image feature extraction device, comprising: an isolated minutia extracting unit that combines the corresponding minutia to obtain an isolated minutia without the corresponding minutia.