KR100348520B1 - Fingerprint Image Correction Method - Google Patents

Abstract

The present invention relates to a method for correcting a fingerprint image, the method comprising: generating a fingerprint image using a touch light emitting device, and calculating a contrast value of the fingerprint image when the fingerprint image is generated in the step of generating the fingerprint image. And dividing the fingerprint image obtained from the touch light emitting device into a plurality of blocks if the contrast value of the fingerprint image is calculated in the step of calculating the contrast value of the fingerprint image. If the fingerprint image is divided into a plurality of blocks, the contrast value calculated in the step of calculating the contrast value of the fingerprint image is compared with the fingerprint image processing apparatus. Correcting blocks that have been determined, and correcting blocks determined to be bad block areas. When the correction is performed, the fingerprint image processing apparatus combines the corrected block area with the normal block area to output a normal image, thereby improving the processing speed of the fingerprint image and improving the accuracy of personal identification using the fingerprint image. have.

Description

Fingerprint Image Correction Method

The present invention relates to a fingerprint image correction method, and in particular, when a person has a wet or dry area due to personal characteristics, a wet or dry area generated in one fingerprint image is corrected to calculate a normal fingerprint image. It relates to a method for correcting a fingerprint image.

With the rapid dissemination of personal computers and communication networks, there is an increasing demand for personal identification procedures such as permits in public offices, access control in certain areas requiring security, commerce using credit cards, and deposit and withdrawal from banks. Personal identity verification uses security methods such as keys, ID cards and passwords. In the case of using a security device such as a key, ID card, password, etc., there is a problem of easy copying or loss.

Recently, biometric security devices have been developed and spread to prevent duplication or loss of security devices. Biometric security devices utilize a variety of biological parts, such as fingerprints, palms, eyes, faces, and voices, and these biometric parts utilize unique physical characteristics of a person. By identifying personal identity using physical characteristics, accidents caused by duplication or loss can be prevented. There is a method using a fingerprint formed on a finger as a method of identifying an individual's identity using physical characteristics.

The identification method of the personal identity using the fingerprint will be described with reference to the accompanying drawings. 1 is a cross-sectional view showing a schematic configuration of a conventional fingerprint recognition device. As shown in the drawing, the contact light emitting element 10, the lens 21, the CCD camera 22, and the image processor 23 are constituted. The touch light emitting device 10 generates an optical image formed on the touched finger 1 when the identity checker's finger 1 is touched while AC is supplied. The touch light emitting device 10 that generates an optical image upon contact of the finger 1 is composed of a transparent electrode layer 11, a light emitting layer 12, and an insulating layer 13.

When an alternating current (AC) is supplied to the transparent electrode layer 11 and the finger 1 contacts the insulating layer 13, a magnetic field is formed between the finger 1 and the transparent electrode layer 11, and the light emitting layer 12 is formed by the magnetic field. ) To generate light. Light emitted from the light emitting layer 12 is generated according to a pattern of ridges and valleys of the fingerprint formed on the finger 1 to form an optical image. The optical image generated in the light emitting layer 12 passes through the transparent electrode layer 11 and the lens 21 and is incident on the charge coupled device camera 22. The lens 21 uses a concave lens to adjust the focus of the optical image incident on the CCD camera 22.

The CCD camera 22 having received the optical image converts the incident optical image into an electrical signal and outputs a fingerprint image. The fingerprint image output from the CCD camera 22 is received by the image processor 23. The image processor 23 outputs the fingerprint image by correcting the received fingerprint image. The fingerprint image output from the image processor 23 is shown in Figs. 2A and 2B.

The fingerprint image shown in FIG. 2A represents a fingerprint image output from the image processor 23 when a plurality of wet or dry areas exist in one fingerprint image area. The fingerprint image shown in FIG. 2B represents one fingerprint image area. The fingerprint image output from the image processing unit 23 is shown when one of the wet and dry areas is wide.

When at least one wet or dry region exists in one fingerprint image, a good fingerprint image may not be obtained. As a result, there is a problem in that it is impossible to accurately extract the singularity of the individual such as the continuity of the ridge and the tendency of the ridge formed on the fingerprint image.

An object of the present invention is to divide a plurality of blocks of a fingerprint image of a person, if a bad block area is generated by wet or dry, corrects the bad block area and combines with the normal bad area to calculate a fingerprint image. To provide a correction method of.

Another object of the present invention is to improve the processing speed of the fingerprint image by correcting the wet or dry region blocked after blocking the wet or dry region that can be generated in one fingerprint image.

Another object of the present invention is to improve the accuracy of identification using fingerprint images by extracting a fingerprint image by correcting the blocked wet or dry region after blocking the wet or dry region that can be generated in one fingerprint image. have.

1 is a cross-sectional view showing a schematic configuration of a conventional fingerprint recognition device,

2A and 2B show an example of a fingerprint having a bad block area;

3 is a block diagram showing a configuration of a fingerprint recognition device for implementing a method for correcting a fingerprint image according to the present invention;

4 is a flowchart showing a method of correcting a fingerprint image according to the present invention;

5A to 5E illustrate an embodiment of a fingerprint image correction method according to the present invention;

6a to 6e are diagrams showing another embodiment of the fingerprint image correction method according to the present invention.

<Code Description of Main Parts of Drawing>

10: contact light emitting element 11: transparent electrode layer

12: light emitting layer 13: insulating layer

21: lens 22: CCD camera

23: image processing unit 31: interface unit

32: inverter driver 33: inverter

34: fingerprint image processing device

In the fingerprint image correction method of the present invention, if the fingerprint image is generated in the step of generating the fingerprint image using the touch light emitting device and in the step of generating the fingerprint image, the fingerprint image processing apparatus calculates the contrast value of the fingerprint image and the fingerprint. When the contrast value of the fingerprint image is calculated in calculating the contrast value of the image, the fingerprint image is divided into a plurality of blocks, and the fingerprint image is divided into a plurality of blocks. When the contrast value is calculated in the step of calculating the contrast value of the fingerprint image when the block is divided, the contrast value calculated by the fingerprint image processing apparatus and the plurality of block areas are compared. Correcting the block determined as the block area and correcting the block determined as the bad block area When the defective area is corrected in the step of performing the fingerprint image processing apparatus characterized in that the combination of the corrected block area and the normal block area to output a normal image.

(Example)

Hereinafter, the present invention will be described with reference to the accompanying drawings.

3 is a block diagram showing a configuration of a fingerprint recognition device for implementing a method for correcting a fingerprint image according to the present invention, and FIG. 4 is a flowchart showing a method for correcting a fingerprint image according to the present invention. As shown, when the fingerprint image is generated in the step S11 to S13 of generating the fingerprint image using the touch light emitting device 10 and the step S11 to S13 of generating the fingerprint image, the fingerprint image processing apparatus 34 In step (S14) of calculating the contrast value of the fingerprint image, and the contrast value of the fingerprint image is calculated in the step (S14) of calculating the contrast value of the fingerprint image a plurality of fingerprint images obtained by the touch light emitting device 10 When the fingerprint image is divided into a plurality of blocks (S15) and the fingerprint image is divided into a plurality of blocks (S15), the contrast value is calculated in the step (S14) of calculating the contrast value of the fingerprint image. Comparing the contrast value calculated by the fingerprint image processing device 34 with the plurality of block areas, if it is larger or smaller than the contrast value, determining the bad block area and correcting the block determined as the bad block area (S16 to S19). )Wow, If the defective area is corrected in steps S16 to S19, the corrected block area is corrected (S16 to S19), and the fingerprint image processing unit 34 combines the corrected block area and the normal block area to output a normal image (S20 to S19). S21).

Referring to the configuration and operation of the present invention in more detail as follows.

A fingerprint reader is used to use a fingerprint formed on a finger as a method of identifying an individual's identity using physical characteristics. The fingerprint recognition device for implementing the present invention is largely composed of the contact light emitting device 10, the image processing unit 23, the inverter 33 and the fingerprint image processing device 34 as shown in FIG. When the contact light emitting device 10 is composed of the transparent electrode layer 11, the light emitting layer 12, and the insulating layer 13, as shown in FIG. 1, the contact light emitting device 10 generates a fingerprint image as an optical signal when the fingerprint contacts the insulating layer 13. (S11 to S13)

In the process of generating a fingerprint image, the touch light emitting device 10 that generates an optical signal is first touched when the finger 1 contacts the insulating layer 13 while AC is supplied to the transparent electrode layer 11. A magnetic field is formed between 1) and the transparent electrode layer 11, and light is emitted from the light emitting layer 12 by this magnetic field (S11). The light emitted from the light emitting layer 12 forms an optical image according to the pattern of the ridges and valleys of the fingerprint formed on the finger 1, and the optical image is a concave lens used to adjust the focus of the transparent electrode layer 11 and the optical image. After passing through the lens 21, it is incident on the CCD camera 22.

The optical image incident on the CCD camera 22 is an optical image generated along the ridge, which is an embossed portion formed on the fingerprint, and the ridge portion is bright in color and the negative portion is dark in the fingerprint in contact with the light emitting element 10. Will appear. The optical image generated by distinguishing the dark portion from the bright portion according to the ridges and valleys of the fingerprint is incident on the CCD camera 22 and then converted into an electrical signal and transmitted to the image processing unit 23.

The image processor 23 stores the fingerprint images according to the resolution of the fingerprint image received from the CCD camera 22 and sequentially transmits the stored fingerprint images to the fingerprint image processing apparatus 34. A computer system is used as the fingerprint image processing apparatus 34. The fingerprint image processing apparatus 34 drives the inverter 33 to receive a fingerprint image from the image processing unit 23, and the driving power is supplied to the touch light emitting device 10. Supply control of AC. In order to drive the inverter 33, the fingerprint image processing apparatus 34 applies a drive control signal to the inverter driver 32 through the interface unit 31.

The inverter driver 32 receives the drive control signal and transmits an inverter drive signal to the inverter 33 according to the received drive control signal, and the inverter 33 is activated by the received inverter drive signal to generate alternating current (AC). The touch light emitting device 10 is transmitted. The touch light emitting device 10 receiving the alternating current (AC) from the inverter 33 generates an optical image when the finger 1 contacts the surface. In order to check whether the finger 1 is in contact, the fingerprint image processing apparatus 34 receives a contact signal output from the CCD camera 22 through the interface unit 31 to check whether the fingerprint is in contact (S12). . When it is confirmed that the fingerprint is in contact with the surface of the touch light emitting device 10, the fingerprint processing apparatus 10 generates a fingerprint image received from the image processing unit 23 (S13).

When the fingerprint image is generated, the fingerprint image processing apparatus 34 calculates a contrast value of the fingerprint image (S14). The contrast value of the fingerprint image is the contrast formed by the ridges and valleys formed on the finger 1, and the ridges, which are embossed portions, appear bright in color and the valleys, which are intaglio portions, in the fingerprint contacting the surface of the light emitting device 10 Appears. When the contrast value of the fingerprint image is calculated by the fingerprint image processing apparatus 34, the fingerprint image obtained by the touch light emitting device 10 is divided into a plurality of blocks (S15). Divide the entire fingerprint image into multiple blocks and determine the local ridge representation in each block area. Here, the ridge representative direction represents the original characteristics of the fingerprint image.

When the fingerprint image is divided into a plurality of blocks, the contrast value calculated in the step S14 of calculating the contrast value of the fingerprint image is compared with the fingerprint image processing device 34, and is determined to be a bad block area when the fingerprint image is larger or smaller than the contrast value. The block determined to be a bad block area is corrected (S16 to S19).

When a bad block area is generated, the process of correcting this first checks whether there is a bad block area in each block area divided into a plurality of blocks (S16). The identification method of the bad block area determines whether the ridge direction of each block area is correctly formed in the fingerprint image processing apparatus 34. If the ridge direction of each block area is not formed correctly, the fingerprint image processing apparatus 34 checks whether the bad block area is a wet area (S17). As a criterion for confirming the wet area, when the image appearing in each block appears in a dark color as a whole, and the ridge direction of the light color part cannot be confirmed, the fingerprint image processing apparatus 34 determines the wet area.

As a result of the check in the fingerprint image processing apparatus 34, if the bad block area is the wet area, the fingerprint image of the wet area is corrected by adjusting the contrast value lower than the contrast value (S18). By lowering the contrast value in the region identified as the bad block region, the clue in the ridge direction can be found to some extent, and thus the ridge in the bad block region can be recovered using the clue in the ridge direction.

If the bad block area is a dry area, the fingerprint value of the dry area is corrected by adjusting the contrast value higher than the contrast value (S19). When the bad block area is identified as a dry area, the bad block area is displayed as a bright color as a whole.To correct this problem, first adjust the contrast value of the bad block area due to dryness higher than the reference contrast value, and then find the clue in the ridge direction. . After that, the ridge of the bad block area due to dryness is recovered using the ridge representative direction.

If there is no bad block area in the process of checking whether there is a bad block area in each divided block area, the fingerprint image processing device 34 outputs a normal fingerprint image (S22), and when the normal fingerprint image is output, the fingerprint image processing device ( In step 34), a process of verifying an identity by matching a pre-stored fingerprint image is performed.

If there is no bad block area in each of the divided block areas, the procedure proceeds with matching the pre-stored fingerprint image to verify the identity. However, if the bad block area is generated in at least one of the plurality of divided blocks, the above-described process is performed. Correction is performed. When the correction of the bad block area is completed, a normal image is output by combining the corrected block area and the normal block area (S20 to S21).

In the process of outputting a normal image by combining the corrected block area and the normal block area, first, when the bad block area is corrected, the fingerprint image processing apparatus 34 combines the corrected block area and the normal block area (S20). When the corrected block area is combined with the normal block area, the combined fingerprint image is output (S21), and when the fingerprint image is output, the fingerprint image processing device 34 matches the generated fingerprint image with a pre-stored fingerprint image for identification. The procedure is to proceed.

After dividing a fingerprint image into a plurality of blocks, an embodiment of checking whether each block has a bad block area and correcting a bad block area will be described with reference to the accompanying drawings. 5A to 5E are roads showing an embodiment of a fingerprint image correction method according to the present invention, and FIG. 5A shows an original image of a fingerprint. When the original image of the fingerprint shown in FIG. 5A is obtained by using the touch light emitting device 10 and then processed by the fingerprint image processing apparatus 34, an image as shown in FIG. 5B is obtained.

When the fingerprint image illustrated in FIG. 5B is confirmed as defective, a defective region such as the fingerprint image illustrated in FIG. 5C may be identified. Here, the defective area is an example in which a predetermined area is displayed in light color by wet. When the defective area shown in FIG. 5C is divided into two defective block areas through a process of dividing a fingerprint image, the two defective block areas are separately divided and then corrected. When two bad block areas are corrected in the fingerprint image shown in FIG. 5D, the corrected bad block area is combined with the normal block area as shown in FIG. 5E to completely correct and output the normal fingerprint image.

Another embodiment of the process of correcting a fingerprint image is shown in FIGS. 6A-6E. 6A is an original image of a fingerprint. When the original image of the fingerprint shown in FIG. 6A is acquired by using the touch light emitting device 10 and then processed by the fingerprint image processing apparatus 34, an image as shown in FIG. 6B is obtained. When the fingerprint image illustrated in FIG. 6B is confirmed as defective, a defective region such as the fingerprint image illustrated in FIG. 6C may be identified.

Here, the case where the defective area is generated in one block is described. In the defective area shown in FIG. 6C, one defective block area is separately divided as in the fingerprint image shown in FIG. 6D and then corrected. When one bad block area shown in FIG. 6D is corrected, the corrected bad block area is combined with the normal block area as shown in FIG. 6E to completely correct and output a normal fingerprint image.

As described above, when there is a wet or dry area in a person's fingerprint, the fingerprint image is divided into normal and abnormal block areas in the divided block, and a bad block area is generated by wet or dry. By correcting the block and calculating the corrected fingerprint image by combining with the normal block area, the processing speed of the fingerprint recognition device can be improved, and more accurate fingerprint image can be obtained, thereby improving the accuracy of identity verification using the fingerprint image. Let's go.

As described above, the present invention divides a fingerprint image into a plurality of blocks, and if a bad block region is generated by wet or dry, corrects the generated block, and then calculates the corrected fingerprint image by combining with the normal block region. It is possible to improve the processing speed of, and to obtain a more accurate fingerprint image, thereby providing an effect of improving the accuracy of identification using the fingerprint image.

Claims (6)

A method of correcting a fingerprint image using a fingerprint image processing apparatus having a touch light emitting device, Generating a fingerprint image using the touch light emitting device; Calculating a contrast value of the fingerprint image in the fingerprint image processing apparatus when the fingerprint image is generated in the step of generating the fingerprint image; Dividing the fingerprint image obtained by the touch light emitting device into a plurality of blocks when the contrast value of the fingerprint image is calculated in calculating the contrast value of the fingerprint image; In the step of dividing the fingerprint image into a plurality of blocks, the brightness value of the fingerprint image is calculated when the fingerprint image is divided into a plurality of blocks. Comparing areas to determine a bad block area if the value is larger or smaller than the contrast value and correcting the block determined to be a bad block area; And If the defective area is corrected in the step of correcting the block determined as the bad block area, the fingerprint image processing apparatus comprises a step of outputting a normal image by combining the corrected block area and the normal block area. Correction method. The method of claim 1, wherein the generating of the fingerprint image comprises: inputting a fingerprint for contacting the fingerprint with the touch light emitting device; Confirming whether the fingerprint is in contact with the touch light emitting device in the step of inputting the fingerprint; And And obtaining a fingerprint image of the contacted fingerprint when the fingerprint is in contact with the touch light emitting device in the step of confirming whether the fingerprint is touched. The method of claim 1, wherein the brightness value of the fingerprint image in the step of calculating the brightness value of the fingerprint image is used as the contrast generated by the ridge and the valley in the fingerprint. The method of claim 1, wherein the determining of the bad block area and correcting the block determined as the bad block area is performed by checking whether there is a bad block area in each block area divided into a plurality of blocks. step; Checking whether a bad block area is a wet area if there is a bad block area in the step of checking whether the bad block area exists; Correcting the fingerprint image of the wet area by adjusting the contrast level to a lower value than the contrast value if the bad block area is a wet area in the step of checking whether the bad block area is a wet area; And In the step of checking whether the wet area is a wet area, if the bad block area is not a wet area, the image is configured to correct the fingerprint image of the dry area by adjusting the contrast value higher than the contrast value. Correction method. 5. The method of claim 4, further comprising outputting a normal fingerprint image if there is no bad block area in the step of checking whether there is a bad block area. 5. The method of claim 4, wherein the outputting of the normal image by combining the corrected block area and the normal block area comprises correcting the corrected block area and normal when correcting the block determined as the bad block area. Combining the block areas; And And outputting a combined fingerprint image when the corrected block region and the normal block region are combined in the combining of the corrected block region and the normal block region.