KR101676000B1 - Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person - Google Patents

Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person Download PDF

Info

Publication number
KR101676000B1
KR101676000B1 KR1020150086486A KR20150086486A KR101676000B1 KR 101676000 B1 KR101676000 B1 KR 101676000B1 KR 1020150086486 A KR1020150086486 A KR 1020150086486A KR 20150086486 A KR20150086486 A KR 20150086486A KR 101676000 B1 KR101676000 B1 KR 101676000B1
Authority
KR
South Korea
Prior art keywords
fingerprint
area
image
candidate
detecting
Prior art date
Application number
KR1020150086486A
Other languages
Korean (ko)
Inventor
백영현
신요식
Original Assignee
주식회사 유니온커뮤니티
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 주식회사 유니온커뮤니티 filed Critical 주식회사 유니온커뮤니티
Priority to KR1020150086486A priority Critical patent/KR101676000B1/en
Application granted granted Critical
Publication of KR101676000B1 publication Critical patent/KR101676000B1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G06T11/60—Editing figures and text; Combining figures or text
    • G06K9/00006—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Processing (AREA)

Abstract

A fingerprint detection and security processing method in a digital document in multiple environments is disclosed. The security processing method of the present invention includes a step of extracting a fingerprint area from an original image including a fingerprint image, and a step of performing security processing by performing one of clipping, masking and mosaicing on the fingerprint area, So that the fingerprint included in the fingerprint can be visually confirmed.

Description

FIELD OF THE INVENTION [0001] The present invention relates to a fingerprint detection and security processing method in a digital document created between a bank, a communication company,

The present invention automatically detects the location and existence of a fingerprint included in a user's handwritten format paper and ID document that are imported into a bank, a communication company, an insurance transaction, etc., and managed as a digital document, A method for detecting fingerprint in a digital document and security processing in a multiple environment that performs a fingerprint detection, a region extraction, a deletion, a mosaic).

Fingerprint is the swelling of sweat glands to form a constant flow, which is different from person to person, and has the characteristic of being unchanged for life as it was when born, and is widely used for personal authentication.

For example, various documents are used during various transaction processes between a bank, a communication company, an insurance company, and an individual. At this time, a personal identification number of a customer or a fingerprint may be taken for personal authentication.

The generated documents were scanned and stored as digital images in formats such as jpg, tiff, bmp, and jpg2000. However, as the need for personal information becomes important and related legislation is implemented, security measures must be taken such as to mask or remove personal information, such as resident registration number and fingerprint, There is a need to do.

In order to secure the resident registration number and the fingerprint recorded in the digital document, the resident number and the location of the fingerprint must be confirmed from the image file and security processing should be performed on the position. The resident registration number can be processed by applying the character recognition technology (OCR). However, there is no separate solution for the fingerprint. At present, only about 1.8 billion cumulative documents are estimated to be processed, and there is a need for a solution that automatically deletes the fingerprint image parts included in these documents.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for automatic fingerprint detection and security processing in a multiple environment capable of automatically detecting the position of a fingerprint existing at an irregular position rather than a specific position in a document by applying image processing and fingerprint recognition technology .

It is still another object of the present invention to provide a method for automatic fingerprint detection and security processing in multiple environments capable of performing security processing such as masking based on the coordinates of an automatically detected fingerprint.

According to another aspect of the present invention, there is provided a security processing method for a digital document, the fingerprint area detection unit extracting a fingerprint area from an original image including a fingerprint image, wherein the fingerprint area is a polygonal area including the fingerprint image And performing security processing by performing one of clipping, masking and mosaicing on the fingerprint area.

According to the embodiment, the step of extracting the fingerprint area may include a candidate area detection step of selecting, as candidate areas, a region in which a plurality of lines having the same directionality are densified with respect to the original image, And a fingerprint area detection step of determining a fingerprint area as an area where the feature image of the fingerprint is detected among the candidate areas.

As one specific method of detecting the candidate region, the candidate region detecting unit may select a region in which one or more regions in which a plurality of lines of a reference length are densely arranged in the same direction are adjacent to each other, as candidate regions. Here, the reference length may be set smaller than the length of the fingerprint area.

As one specific method of detecting the fingerprint area, the fingerprint area detector may determine the candidate area as a fingerprint area when the minutia value obtained by applying the fingerprint recognition algorithm to the candidate area is equal to or greater than a reference value.

Wherein the fingerprint area detecting unit detects the candidate area as an elliptical shape when the length ratio of the major axis to the minor axis of the ellipse is within a preset effective range, Area can be judged.

According to another embodiment, the security processing method of the present invention may further include a preprocessing step of removing noise included in the original image before the preprocessing unit performs the candidate region detection step. Further, if the original image is a color image, the preprocessing unit may further perform a preprocessing step of converting the original image into a gray-level image before performing the candidate region detection step.

According to the method of the present invention, the fingerprint area can be detected from the image document including the fingerprint image, and the personal information protection process can be performed by cutting, masking, or mosaicing the fingerprint so that the fingerprint can not be visually confirmed.

1 is a block diagram of a fingerprint processing apparatus of the present invention,
2 is a flowchart provided in the description of the fingerprint detection and security processing method in the digital document of the present invention, and
FIG. 3 is a diagram showing an example in which the method of FIG. 2 is applied.

Referring to FIG. 1, the security processing apparatus 100 of the present invention identifies a fingerprint image from personal information contained in a digital image document stored in a file form, and identifies the fingerprint image area (hereinafter, referred to as a 'fingerprint area' ) Is not visible to a third party.

Here, the digital image document is a document (i.e., image) stored in a digital image format such as jpg, tiff, bmp, or jpg2000, and a fingerprint image may be included in the image. Mainly, paper documents can be scanned or created, or various word processor programs can be created and saved as images, or they can be created as images directly by an image production program. In addition, the digital image document may be composed of a plurality of pages. However, the following description will be based on a page image (or a document thereof) and is simply referred to as an 'original document' or an 'original image'.

As in Figs. 3A and 3F, the fingerprint area may be a polygonal area including a fingerprint image, and the minimum square that may include a fingerprint image does not infringe another area.

In addition, the security processing for the fingerprint area means that a third party can erase or mask the fingerprint image included in the digital original document so that the fingerprint image can not be visually confirmed. (1) (2) masking or mosaicing the fingerprint area opaque with a specific color, and the like.

The security processing apparatus 100 of the present invention can be operated in a stand-alone manner and may be connected to a database (not shown) in which a digital document to be subjected to security processing is stored via a network . Hereinafter, the security processing apparatus 100 that operates alone will be mainly described.

1, the security processing apparatus 100 includes a storage medium 101 storing various digital original documents to be subjected to security processing, a memory 103, a controller 110 for performing fingerprint detection and security processing ).

The control unit 110 further includes a preprocessing unit 111, a candidate region detection unit 113, a fingerprint region detection unit 115, and a security processing unit 117. Hereinafter, the operation of each component of the controller 110 will be described while referring to the fingerprint detection and security processing method of the present invention with reference to FIG. 2 and FIG. As described above, if one document file includes a document (image) of a plurality of pages, the security processing method shown in Fig. 2 is performed on a page-by-page basis.

<1. Fingerprint area search>

&Lt; Convert color image to gray image: S201, S203 >

If the original document loaded from the storage medium 101 into the memory 103 is a color document (S203), the preprocessing unit 111 removes the color from the original document and displays the gray Into a document (S203).

However, since the method of the present invention can be performed even in the state of a color image, conversion to a gray image is not an essential step. However, since the fingerprint area can be detected even in the gray image state, the color image does not need to be a color image. In the case of the color image, the file size is relatively large, .

The preprocessing unit 111 can convert a color image document into a gray level image document using a color space model capable of extracting a gray level image among color space models (YCbCr, HSI, CMYK, etc.) have.

<Noise removal in document: S205>

Since the digital document stored in the storage medium 101 is often a document that scans a paper document and such a scanned document contains considerable noise, it is necessary to distinguish and enhance the background and the foreground (letter, line, image, etc.) have. Also, when the noise is removed, the amount of data to be processed is reduced, so it is preferable to perform the pre-processing step.

Various noise reduction techniques are possible, but multichannel sampling techniques, such as wavelet transforms and discrete Fourier transforms, are useful in clarifying the background and clarifying the shape of individual foregrounds.

For example, wavelet transforms can be made up of a set of special signals that model a signal, system, or process. This special signal is called a wavelet, and it is expressed as a random waveform by scaling a wavelet of a certain locally existing one, or by scaling the scaling. The two-dimensional wavelet transform performs the low-pass filter and the high-pass filter alternately. The wavelet transform obtains four frequency bands (LL, LH, HL, HH). The wavelet transform can be analyzed in detail and has the advantage of improving data processing with noise reduction and high compression ratio.

<Candidate region detection step: S207>

A step of detecting a fingerprint area is performed on the gray level document from which noise has been removed. A variety of methods for fingerprint area detection can be presented. In the present invention, the fingerprint area detection is performed in two steps, paying attention to the features of the fingerprint. In the first step, considering that the fingerprint is a set of a plurality of lines having substantially the same direction, the candidate region detecting unit 113 first selects candidate regions in which a plurality of lines having the same directionality are concentrated. In the second step, the fingerprint area detecting unit 115 detects a fingerprint area in the corresponding document by judging whether or not the candidate areas have a characteristic inherent to the fingerprint.

The candidate region detection unit 113 performs a candidate region detection algorithm on the entire document to find a candidate region. Various algorithms for detecting a candidate region in which a plurality of lines having the same directionality are dense can be used.

For example, individual lines forming a fingerprint have a spiral shape, but a set of segmented lines of the spiral can be perceived in the same direction in a certain length unit. Therefore, it is possible to determine that a region in which a plurality of lines of the reference length are densely arranged in the same direction as one or more adjacent regions is regarded as a candidate region. Here, the reference length corresponds to the segmented unit segment of the helix forming the fingerprint, and therefore should be set to be smaller than the typical horizontal or vertical length of the fingerprint area.

3 is presented in a color image state to facilitate understanding, but is preferably processed as a gray level image. 3 (a) is an original scanned image document to be subjected to security processing, in which characters and some lines are displayed, and fingerprints are printed on one side. (b) and (c) illustrate an object area in which a candidate region detection algorithm is to be performed, for convenience of explanation, and (d) and (e) FIG. 8 is a diagram illustrating a result of performing a detection algorithm. FIG.

The candidate region detection unit 113 does not judge that the candidate region is the candidate region, while the candidate region judges the (b) region. (b) is a pattern in which a plurality of regions a1, a2, a3, and a4 in which a plurality of lines having a reference length or more are densely arranged in the same direction are disposed adjacent to each other. On the other hand, (c) does not show a region in which a plurality of lines having a reference length or more are concentrated in the same direction.

&Lt; Final detection of fingerprint area: S209 >

When the candidate region detection unit 113 finds the candidate region, the fingerprint region detection unit 115 determines whether the candidate region is a fingerprint by judging whether or not the candidate regions have the fingerprint characteristic.

Here, the fingerprint characteristic is a ridge and a valley having a repetitive line shape that a fingerprint must have, a bifurcation at which a ridge breaks, an end point at which a ridge is broken, and a center point where a bend is the largest at the top. In addition, there are also 'lower center point' where the curvature is the largest at the lower side of the fingerprint characteristic, and a delta where the ridge flow of the fingerprint is gathered from three directions.

Therefore, the fingerprint area detecting unit 115 can determine whether the candidate area includes the characteristic of the fingerprint and finally determine whether the candidate area is the fingerprint area. At this time, one of the following processes can be selectively used according to convenience and extraction difficulty.

(Option 1) Fingerprint Pattern It is possible to determine whether the candidate region is a fingerprint image by applying a classification algorithm according to the type of the fingerprint through the pattern search. A method of determining whether a ellipse having a fingerprint characteristically exists in a candidate region and determining whether a length ratio (so-called centrifugal ratio) of the major axis and minor axis of the ellipse is within a preset effective range, It can be judged as the final fingerprint area.

(Option 2) It is possible to determine whether the candidate region is a fingerprint image by applying a unique feature point extraction method used for fingerprint recognition. By applying the technique of extracting minutia data for fingerprint recognition, minutiae extraction is performed by binarization, sessionization, etc., and if the minutia value obtained is more than a certain score, it can be judged as the final fingerprint area.

FIG. 3 (b) shows an image of the actual fingerprint, and thus has all the features of the fingerprint. Accordingly, FIG. 3 (b) can be finally determined as the fingerprint area by the fingerprint area detecting unit 115 even if any of the above three methods is applied.

According to the embodiment, the fingerprint area detecting unit 115 detects a region that is not determined as a fingerprint area in one page of the document but is determined to be very similar to the fingerprint in the final judgment, Tags can be given or stored in a specific folder.

<2. Security processing for the fingerprint area: S211>

The fingerprint area detection unit 115 finally selects a fingerprint area, extracts the coordinates of the fingerprint area, and provides the extracted coordinates to the security processing unit 117. The security processing unit 117 performs security processing based on the coordinates extracted by the fingerprint area detecting unit 115.

The security processing performed by the security processing unit 117 is to prevent the fingerprint from being visually recognized in the document. Accordingly, as described above, the security processing unit 117 may perform 'clipping' the corresponding coordinate area on the image processing technique, mosaic the corresponding area, or perform masking as shown in FIG. 3 (f).

The security processing unit 117 may additionally perform a procedure such as storing the document with security processing in different folders according to a predetermined classification algorithm or assigning another tag. For example, when one fingerprint area is detected and secured in one page and two or more fingerprints are recognized and secured in one page, the security processing unit 117 may not be able to detect the fingerprint area in one page, When a similar area is detected, or when other exception occurs, the secured document can be sorted.

Claims (8)

A method for security processing of a digital document generated between a bank, a communication company or an insurance company and an individual,
Extracting a fingerprint area from an original image of the digital document including a fingerprint image; And
And performing security processing on the digital document by performing one of clipping, masking, and mosaicing on the fingerprint area,
Wherein the fingerprint area is a polygonal area including the fingerprint image,
Wherein the step of extracting the fingerprint area comprises:
A candidate region detecting step of detecting a region in which a plurality of lines having the same directionality are densified as a candidate region with respect to the original image; And
And a fingerprint area detecting step of detecting a feature image of the fingerprint in the candidate area as a fingerprint area,
Wherein the candidate region detecting unit selects the region in which the plurality of lines of the reference length are densely arranged in the same direction as one or more adjacent regions as the candidate region, and the reference length is the segmented region of the helix forming the fingerprint Corresponds to the length of the unit line segment,
Wherein the fingerprint area detecting unit determines the candidate area as a fingerprint area when the candidate area is in an elliptic shape and the length ratio between the major axis and the minor axis of the ellipse is within a preset effective range. A method for security processing of a digital document.
delete delete The method according to claim 1,
In the fingerprint area detecting step,
Wherein the fingerprint area detecting unit determines the candidate area as a fingerprint area when the minutiae point value obtained by applying the fingerprint recognition algorithm to the candidate area is equal to or greater than a reference value.
delete delete The method according to claim 1,
Further comprising a preprocessing step of removing noise included in the original image before the preprocessing unit performs the candidate region detection step.
The method according to claim 1,
Further comprising a pre-processing step of, when the original image is a color image, converting the original image into a gray-level image before the pre-processing performing unit performs the candidate area detecting step.
KR1020150086486A 2015-06-18 2015-06-18 Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person KR101676000B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020150086486A KR101676000B1 (en) 2015-06-18 2015-06-18 Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020150086486A KR101676000B1 (en) 2015-06-18 2015-06-18 Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person

Publications (1)

Publication Number Publication Date
KR101676000B1 true KR101676000B1 (en) 2016-11-14

Family

ID=57528514

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020150086486A KR101676000B1 (en) 2015-06-18 2015-06-18 Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person

Country Status (1)

Country Link
KR (1) KR101676000B1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108235814A (en) * 2018-01-05 2018-06-29 深圳市汇顶科技股份有限公司 The method and terminal device of user interface interaction
KR20190096116A (en) 2018-02-08 2019-08-19 한화테크윈 주식회사 Security method for biometric information
WO2020197048A1 (en) * 2019-03-28 2020-10-01 Samsung Electronics Co., Ltd. Electronic device and method for securing personal information included in image
CN112435688A (en) * 2020-11-20 2021-03-02 腾讯音乐娱乐科技（深圳）有限公司 Audio recognition method, server and storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100267262B1 (en) * 1998-03-13 2000-11-01 김영진 Method for extracting fingerprint features
JP2007251929A (en) * 2006-01-31 2007-09-27 Konica Minolta Systems Lab Inc Selective image encoding and replacement
KR101460964B1 (en) * 2013-05-14 2014-11-12 주식회사 슈프리마 Apparatus and method for capturing fingerprint image

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100267262B1 (en) * 1998-03-13 2000-11-01 김영진 Method for extracting fingerprint features
JP2007251929A (en) * 2006-01-31 2007-09-27 Konica Minolta Systems Lab Inc Selective image encoding and replacement
KR101460964B1 (en) * 2013-05-14 2014-11-12 주식회사 슈프리마 Apparatus and method for capturing fingerprint image

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
지문 영상의 특징 정보 추출을 위한 효율적인 주름선 추출 방법, 전자공학회논문지 IE편 제44권2호 (2007)* *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108235814A (en) * 2018-01-05 2018-06-29 深圳市汇顶科技股份有限公司 The method and terminal device of user interface interaction
KR20190096116A (en) 2018-02-08 2019-08-19 한화테크윈 주식회사 Security method for biometric information
WO2020197048A1 (en) * 2019-03-28 2020-10-01 Samsung Electronics Co., Ltd. Electronic device and method for securing personal information included in image
US11392719B2 (en) 2019-03-28 2022-07-19 Samsung Electronics Co., Ltd. Electronic device and method for securing personal information included in image
CN112435688A (en) * 2020-11-20 2021-03-02 腾讯音乐娱乐科技（深圳）有限公司 Audio recognition method, server and storage medium

Similar Documents

Publication Publication Date Title
US11302109B2 (en) Range and/or polarity-based thresholding for improved data extraction
Shi et al. A steerable directional local profile technique for extraction of handwritten arabic text lines
EP2783328B1 (en) Text detection using multi-layer connected components with histograms
US8594431B2 (en) Adaptive partial character recognition
US7062099B2 (en) Image processing method and apparatus using self-adaptive binarization
Farahmand et al. Document image noises and removal methods
JP6080259B2 (en) Character cutting device and character cutting method
EP3002712A2 (en) Horizontal and vertical line detection and removal for document images
KR101676000B1 (en) Method for Detecting and Security-Processing Fingerprint in Digital Documents made between Bank, Telecommunications Firm or Insurance Company and Private person
EP2104059B1 (en) Apparatus and method for processing image
US8014574B2 (en) Character noise eliminating apparatus, character noise eliminating method, and character noise eliminating program
JP4821663B2 (en) Character noise elimination device, character noise elimination method, character noise elimination program
US20160292528A1 (en) Character segmentation device, character recognition device and character segmentation method
US10885326B2 (en) Character recognition method
JP6599672B2 (en) Character cutout device, character recognition device, and character cutout method
JP4904330B2 (en) Method and apparatus for extracting text from an image
JP2011248702A (en) Image processing device, image processing method, image processing program, and program storage medium
CN111209865A (en) File content extraction method and device, electronic equipment and storage medium
CN110321887B (en) Document image processing method, document image processing apparatus, and storage medium
Zhan et al. A robust split-and-merge text segmentation approach for images
KR100726473B1 (en) Apparatus for classifying an image and method therefor
JP6554352B2 (en) Form discrimination program, form discrimination method, and form discrimination apparatus
CN113033562A (en) Image processing method, device, equipment and storage medium
Hidayatullah et al. License plate detection and recognition for Indonesian cars
Ghiani et al. On the interoperability of capture devices in fingerprint presentation attacks detection

Legal Events

Date Code Title Description
E701 Decision to grant or registration of patent right
GRNT Written decision to grant