JPS5924384A - Checking circuit of feature point - Google Patents

Checking circuit of feature point

Info

Publication number
JPS5924384A
JPS5924384A JP57133230A JP13323082A JPS5924384A JP S5924384 A JPS5924384 A JP S5924384A JP 57133230 A JP57133230 A JP 57133230A JP 13323082 A JP13323082 A JP 13323082A JP S5924384 A JPS5924384 A JP S5924384A
Authority
JP
Japan
Prior art keywords
feature points
points
point
coordinates
pair
Prior art date
Legal status (The legal status 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 status listed.)
Pending
Application number
JP57133230A
Other languages
Japanese (ja)
Inventor
Hiroshi Asai
淺井 紘
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEC Corp
Original Assignee
NEC Corp
Nippon Electric Co Ltd
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 NEC Corp, Nippon Electric Co Ltd filed Critical NEC Corp
Priority to JP57133230A priority Critical patent/JPS5924384A/en
Publication of JPS5924384A publication Critical patent/JPS5924384A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12Fingerprints or palmprints
    • G06V40/1347Preprocessing; Feature extraction

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Collating Specific Patterns (AREA)

Abstract

PURPOSE:To decide the coincidence between a pair of feature points, by obtaining the coordinates of a feature point near the original point defined by either one of the paired feature points from two pair of memories which hold the data on the feature points of finger prints and then comparing the difference between coordinate values with a threshold value. CONSTITUTION:Coordinates X and Y and direction coefficients Ds and Df are obtained from 2 pair of memories holding the data on a stripe-shaped pattern such as a fingerprint, etc. for the feature points of each pair, e.g., branch points Ms/Mf and Ms'/Mf'. Then the coordinates and direction coefficients of the approximates feature points Ms' and Mf' are obtained from above-mentioned data and by converting those feature points into the coordinates including the feature points Ms and Mf used as original points respectively. Then the difference between new coordinates and direction coefficients of points Ms' and Mf' are obtained and compared with the threshold value set previously. Then it is decided that the point Ms' is coincident with the point Mf' if the above-mentioned difference is less than the threshold value. This decision is carried out with all feature points. Thus the vicinity of the feature points is decided accurately, and a recognizing factor is improved. This can be applied to the judgement of a seal.

Description

【発明の詳細な説明】 不発明は指紋等の縞状のパターンをその%徴を用いて照
合する装置に関するものである。
DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus for verifying a striped pattern such as a fingerprint using its percent signature.

従来、指紋照合をパターン認識技術を用いて行なう方式
は、その紋様の特徴点を用いる方式が偉業されている。
Conventionally, the most successful method of performing fingerprint verification using pattern recognition technology is a method that uses the feature points of the pattern.

即ち、第1図を参照しで、押捺印象画01に現れた指紋
紋様を構成する隆線02 の端点03及び分岐点o4を
特徴点とする0 各特徴点M7は、押捺印象画成は紋様
にょっ゛C固有1こ定義された座標系によって、その位
[(Xf、Y7)、方向Df7Jびその他の特徴組によ
って記述される。
That is, referring to FIG. 1, each feature point M7 is defined as the end point 03 and branch point o4 of the ridge line 02 forming the fingerprint pattern appearing on the stamp impression image 01. By the coordinate system defined by Nyoc C, the position [(Xf, Y7), direction Df7J, and other feature sets are described.

照合原理は、上記’riJ債点の配置の合同性によって
行われるものである。
The matching principle is based on the congruence of the arrangement of the 'riJ bond points.

特徴の抽出手段は特組昭54−39648号明細書(特
開昭56−138174号公報)、また特徴?ζよる照
合手段?こついては特願昭54−98966号明細書(
特開昭56−24675号公叩)等でその実現が示され
ている。♂ころで、照合の過程に於いて照合Jべき探索
指紋と照合されるべきファイルとし−C蓄積されている
ファイル指紋の1つとを照合する場合に、その基本とな
るのは、対応する特徴点「対」を検出し、その過不足が
ないことを検査することである。この特徴点「対」の検
出に於いては、その位置、方向の一致性を検査する前に
、各特徴点「対」候補に対して、それぞれの特徴点即ち
探索指紋の特徴点Mf(、?’=1〜F)とファイル指
紋の特徴点M6(s=1〜S)の互の近傍の特徴点の配
位関係を検査することが非常に重要である。
Feature extraction means are described in Tokugumi No. 54-39648 (Japanese Unexamined Patent Publication No. 56-138174), and feature? Verification method using ζ? Regarding this issue, see the specification of Japanese Patent Application No. 54-98966 (
The realization of this is shown in Japanese Patent Application Laid-Open No. 56-24675). By the way, in the process of matching, when matching the search fingerprint to be matched with one of the stored file fingerprints as the file to be matched, the basic method is to find the corresponding minutiae. The purpose is to detect "pairs" and check that there are no excesses or deficiencies. In detecting this minutiae point "pair", before checking the consistency of their positions and directions, for each minutiae "pair" candidate, each minutiae point, that is, the minutiae point Mf(, It is very important to examine the coordination relationship between the mutually neighboring minutiae of the file fingerprint minutiae M6 (s=1 to S) and the minutiae point M6 (s=1 to S) of the file fingerprint.

ところが第2図(al 、 (b)を参照して同図(8
)に示される端点及び同図(b)に示される分岐点のよ
うに、押捺のインク濃度や、印象の光電変換走査時の性
能に起因して点線で示す如く、特徴点の検出位置の安定
性は隆線の方向に不安定であり、隆線方向と直角方向に
は比較的安定である事実がある。
However, referring to Figure 2 (al, (b)), the same figure (8
) and the branch point shown in (b) of the same figure, due to the ink density of the impression and the performance during photoelectric conversion scanning of the impression, the detection position of the feature point is stabilized as shown by the dotted line. There is a fact that the property is unstable in the direction of the ridge, but relatively stable in the direction perpendicular to the ridge direction.

本発明の目的は、指紋の特徴点「対」検査の際近傍特徴
点配位を検査するための特徴点抽出時の特性を考慮した
より精密な検査方法を提供することにある。
SUMMARY OF THE INVENTION An object of the present invention is to provide a more precise inspection method that takes into consideration the characteristics at the time of minutiae extraction in order to inspect the configuration of neighboring minutiae when inspecting minutiae pairs of fingerprints.

本発明によれば、これまでの検出手段に僅かな変更を付
加するのみで特徴点抽出時の抽出位置変位を効果的に解
決することができる。
According to the present invention, it is possible to effectively solve the problem of extraction position displacement during feature point extraction by adding a slight change to the conventional detection means.

次に図面を参照しながら、本f6EtlJの一実施例屹
ついて詳細に説明する。
Next, one embodiment of this f6EtlJ will be described in detail with reference to the drawings.

第3図は、本発明になる回路の一実施例を示すブロック
図であって、探索指紋の特徴点MS(s=1〜S)を格
能するメモリlo1 ファイル指紋の特徴点Mf(f=
x=F)を格納するメモ1J20.座標変換回路30.
40、検査回路5Q、及び制御部6oから構成される。
FIG. 3 is a block diagram showing an embodiment of the circuit according to the present invention, in which the memory lo1 serves the minutiae MS (s=1 to S) of the search fingerprint; the minutiae Mf (f=
x=F) memo 1J20. Coordinate conversion circuit 30.
40, a test circuit 5Q, and a control section 6o.

メモリ川、 20は制御部からの特徴点番号S及びfを
アドレス信号601.602iCよt)%像点Ms、M
fのデータ(Xs、 Ys、Ds)及び(Xf。
Memory river 20 is the feature point number S and f from the control unit as the address signal 601.602iCt)% image point Ms, M
Data of f (Xs, Ys, Ds) and (Xf.

Y7. Dl )それぞれ出力信号010x、、、(1
及び200X* p+ dに出力する通常の記憶回路で
ある。
Y7. Dl ) respectively output signals 010x, , (1
and 200X*p+d.

才な座標変換回路30.40は同−構造である。The efficient coordinate transformation circuits 30 and 40 have the same structure.

第4図は座標変換回路3oの詳細を示すブロック図であ
る。同図に於いて、座標変換回路3oはレジスタ31z
 * y * a s * N器32X+7+d1 三
角定数ROM33、乗算器34.35.36137及び
加算器33x1減算器33.によって構成される。先に
メモリIOから読出されたデータ(Xs+ Ys+ D
s)は、制御部60からのセット信号603によってレ
ジスタ31x、y、d7こ保持される。 次に%像点M
3の近傍特徴点M s ’が、制御部60のアドレス信
号601によってメモリ10から出力100x、y、d
ζこ出力されるさ、座標変換回路30はレジスタ31x
FIG. 4 is a block diagram showing details of the coordinate conversion circuit 3o. In the figure, the coordinate conversion circuit 3o is a register 31z.
* y * a s * N unit 32 Consisted of. Data previously read from memory IO (Xs+ Ys+ D
s) is held in the registers 31x, y, and d7 by a set signal 603 from the control unit 60. Next, % image point M
The neighboring feature point M s ' of 3 is output 100x, y, d from the memory 10 by the address signal 601 of the control unit 60.
The coordinate conversion circuit 30 outputs ζ from the register 31x.
.

yId  に保持されている(X、、Y、、DB )と
減算器32X、yIdで減算を行うとともに、D、が三
角定数ROM3B:に供給されμsDs及びsin l
) sの値が信号33c+ 33sに出力され、34.
35.36.37の乗算器と加減算器38x、38yに
よってXs’ :=(Xs’−Xs )asDs+(Y
s’+) s )Sinl)s△。
The subtractor 32X and yId perform subtraction with (X,, Y,, DB) held in yId, and D is supplied to the trigonometric constant ROM 3B: μsDs and sin l
) the value of s is output on signal 33c+33s, 34.
35.36.37 multiplier and adder/subtractor 38x, 38y produce Xs':=(Xs'-Xs)asDs+(Y
s'+) s)Sinl)s△.

YsA=(Ys′−Ys )eosl)s−(Xs’−
Xs )slnDsなる出力X s * ’j s’が
 300x、 300y に出力される。一方300d
ζこは D s =D s −I) s となり、(Xs* YssDs)を原点とする(Xs’
YsA=(Ys'-Ys)eosl)s-(Xs'-
The output X s *'j s' of slnDs is output at 300x and 300y. On the other hand, 300d
ζ is Ds = Ds - I)s, and (Xs'
.

Y?、Ds’)の局所座標変換が行われる0座標変換回
路40の動作は同様であるので説明を省略する。
Y? .

このようにして探索指紋特徴点Msの近傍特徴点M、′
と、ファイル指紋特徴点Mfの近傍特徴点M ftとが
、それぞれ特徴点Ms、M1を原点さする局所△ Ds’ )及び< x、′、 y 7 、 n 、/ 
)がそれぞれ300x。
In this way, the neighboring minutiae M,′ of the search fingerprint minutiae Ms
and the neighboring minutiae point M ft of the file fingerprint minutiae Mf are local ΔDs') whose origin is the minutiae point Ms and M1, respectively, and <x,', y 7 , n , /
) are 300x each.

yId及び400x+3’edに出力されると、これら
を入力として検査回路50は次のような検査処理を行な
う。
When output to yId and 400x+3'ed, the inspection circuit 50 uses these as input and performs the following inspection process.

第5図は検査回路50の詳細を示すブロック図である。FIG. 5 is a block diagram showing details of the inspection circuit 50.

絶対1■減算回路51x、 7* d %閾値発生器5
3、比較器”’X+7e! 及びANDゲート55から
構成される。検査回路50の動作は、信号400x、 
y、 、3と300x、、、d とを絶対値減算を行い
、その出力と、閾値発生器53の出力sax、 ye 
dによる閾値(Tx、Ty+Td)lごよって比較器”
’X+y+dによって △、△、    △7△ lX5−Xfl≦Tx、 IYs−Y715Ty111
)s DfI≦Tdなる判定を行う03つの判定が全て
真であるときANDゲート55は、一致信号500を制
御部60に出力するこさとなる。
Absolute 1■ Subtraction circuit 51x, 7*d% threshold generator 5
3. It is composed of a comparator "'X+7e!" and an AND gate 55. The operation of the test circuit 50 is based on the signals 400x,
Absolute value subtraction is performed on y, , 3 and 300x, , d, and the output is the output of the threshold generator 53, sax, ye
Threshold value by d (Tx, Ty + Td) Comparator by l
'X+y+d △, △, △7△ lX5-Xfl≦Tx, IYs-Y715Ty111
)s DfI≦Td When all three judgments are true, the AND gate 55 outputs a coincidence signal 500 to the control unit 60.

ここで理解を助けるために、上記動作を指紋特徴上で第
6図(a)、(b)、 (C)を用いて説明しよう。探
索指紋特徴点Msと近傍%像点Ms’は(a)、ファイ
ル指紋IPij像点Mfと近傍特徴点Mf′は(b)に
それぞれ図のように配位しているとする0 座標変換回路30及び40によって局所座標変換を行っ
たのは図の局所座標71.72によっそそれぞれ近傍%
像点Ms’、M7’を表現することになる。
In order to facilitate understanding, the above operation will be explained on fingerprint features using FIGS. 6(a), (b), and (C). It is assumed that the search fingerprint minutiae Ms and the neighboring % image point Ms' are arranged as shown in (a), and the file fingerprint IPij image point Mf and the neighboring minutiae point Mf' are arranged as shown in the figure (b).0 Coordinate conversion circuit The local coordinate transformation using 30 and 40 is based on the local coordinates 71 and 72 in the figure, respectively.
Image points Ms' and M7' will be expressed.

これらを相対関係を杉・肴回路50で検査するこ吉は、
第6図(c)で示した局所座標系73によって近傍特徴
点λis1Mfを調べるととJこなる0 このきき、隆
線方向は、特徴点Ms及びM7の方向り、。
Kokichi examines the relative relationships between these using the cedar-appetizer circuit 50.
When we examine the nearby feature point λis1Mf using the local coordinate system 73 shown in FIG. 6(c), we find that the ridge direction is the direction of the feature points Ms and M7.

Dfによって局所座標71.72が定められているので
第6図(C)に於いては局所座標73のy方向に他なら
ない。したがって%微志抽出時の特徴点変位が隆線方向
に起きやすいことを考慮するためには閾値Txs Ty
は74のような一般的な等しい閾値でなく75のような
1′(l1lyとすることによって可能になる0このこ
とは第5図のもβ値発生器53Tx、TyをTx’)l
’yと設定づ−ることにより容易に実現できる0 以上で、本発明の一実施例について説ψjを終えた。制
御部60については既述の動作説明によりこれを構成J
ることは当業者には容易になし得るもので説明を要しな
い0また実施例で示した各回路は、その機能を実現でき
ることを示した一実施例であって、マイクロプロセッサ
等によっても実現出来るものであって、Cれに1我るも
のではない。
Since the local coordinates 71 and 72 are determined by Df, in FIG. 6(C), this is nothing but the y direction of the local coordinates 73. Therefore, in order to take into account that feature point displacement during % fine point extraction tends to occur in the ridge direction, the threshold Txs Ty
is made possible by setting 1' (l1ly) to 1' (l1ly) such as 75 instead of a common equal threshold value such as 74.
0, which can be easily realized by setting ``y.'' This concludes the explanation of one embodiment of the present invention ψj. The control unit 60 is configured according to the operation description given above.
Those skilled in the art can easily do this and no explanation is required.The circuits shown in the examples are examples showing that the functions can be realized, and can also be realized by a microprocessor or the like. It is a thing, and it is not something that belongs to me.

【図面の簡単な説明】[Brief explanation of the drawing]

第1図は指紋紋様の構成と骨撃点を説明する図、第2図
(at、(1))は’F1点抽出時の変位を説明する図
、第3図は本発明になる%像点検査回路の一実施例を示
すフロ・ツク図、第4図は座標変換回路30.40の一
実施911を示1ブロック図、第5図は扶育回路50の
一実施例を示すブロック図、第6図(a) + (b)
 *(clは本発明の機能動作を説明するための図であ
る。 図1c於いて、探索指紋の特徴点メモリ川、ファイル指
紋の特徴点メモリ20、座標変換回路30.40、検査
回路50、及び制御部60をそれぞれ示すものである。 第 1 図 $Z図 ((J)                (b)郷3
図 罎4図 第5回 5Uυ
Figure 1 is a diagram explaining the structure of the fingerprint pattern and the bone shot point, Figure 2 (at, (1)) is a diagram explaining the displacement when extracting the 'F1 point, and Figure 3 is the % image according to the present invention. A flow diagram showing one embodiment of the point inspection circuit, FIG. 4 is a block diagram showing one implementation 911 of the coordinate conversion circuit 30.40, FIG. 5 is a block diagram showing one embodiment of the support circuit 50, Figure 6 (a) + (b)
*(cl is a diagram for explaining the functional operation of the present invention. In FIG. 1c, a search fingerprint minutiae memory 20, a file fingerprint minutiae memory 20, a coordinate conversion circuit 30, 40, an inspection circuit 50, and the control unit 60. Fig. 1 $Z diagram ((J) (b) Go 3
Figure 4 Figure 5th 5Uυ

Claims (1)

【特許請求の範囲】 指紋等の縞状パターンをその%像点で照合する番 装置に於いて、照合すべき2つのパターンの特徴点群を
保持する2組のメモリと、その出力である特徴点データ
を座標変換するための座標変換回路と、閾値検査をする
ための横書回路及び制御部とからなり、上記2組のメモ
リから読出され、た特徴点「対」に対して該叫像点「対
」の各々の近傍特徴点を前記特徴点「対」を原点とする
局D[座標に変換した後、前記近傍特徴府の配置を検査
する際、前記特徴点「対」の降り方向を前記検査の閾値
に反映させることにより、前記縞パターンでの特徴点の
抽出時に発生し易い位置変位に対して効果的に対処する
ことを特徴とする特徴点検査回路。
[Claims] In a device that matches striped patterns such as fingerprints using their % image points, there are two sets of memories that hold feature point groups of two patterns to be matched, and features that are the output thereof. It consists of a coordinate transformation circuit for coordinate transformation of point data, a horizontal writing circuit for threshold inspection, and a control unit. After converting the neighboring feature points of each of the "pairs" of points into coordinates of the station D with the "pair" of feature points as the origin, when inspecting the arrangement of the neighboring feature points, the direction of descent of the "pair" of feature points is A feature point inspection circuit characterized in that by reflecting this in the inspection threshold value, it is possible to effectively deal with positional displacement that is likely to occur when extracting feature points in the striped pattern.
JP57133230A 1982-07-30 1982-07-30 Checking circuit of feature point Pending JPS5924384A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP57133230A JPS5924384A (en) 1982-07-30 1982-07-30 Checking circuit of feature point

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP57133230A JPS5924384A (en) 1982-07-30 1982-07-30 Checking circuit of feature point

Publications (1)

Publication Number Publication Date
JPS5924384A true JPS5924384A (en) 1984-02-08

Family

ID=15099762

Family Applications (1)

Application Number Title Priority Date Filing Date
JP57133230A Pending JPS5924384A (en) 1982-07-30 1982-07-30 Checking circuit of feature point

Country Status (1)

Country Link
JP (1) JPS5924384A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6255776A (en) * 1985-09-04 1987-03-11 Hitachi Ltd Individual discriminating system by fingerprint
US4896363A (en) * 1987-05-28 1990-01-23 Thumbscan, Inc. Apparatus and method for matching image characteristics such as fingerprint minutiae
US5366898A (en) * 1992-03-27 1994-11-22 Dexsil Corporation Method for quantitative determination of total base or acid number of oil
US7043083B2 (en) 2001-03-28 2006-05-09 Nec Corporation Pattern-collating device, pattern-collating method and pattern-collating program
GB2487377B (en) * 2011-01-18 2018-02-14 Aptina Imaging Corp Matching interest points

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5624675A (en) * 1979-08-02 1981-03-09 Nec Corp Fingerprint reference device

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5624675A (en) * 1979-08-02 1981-03-09 Nec Corp Fingerprint reference device

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6255776A (en) * 1985-09-04 1987-03-11 Hitachi Ltd Individual discriminating system by fingerprint
US4896363A (en) * 1987-05-28 1990-01-23 Thumbscan, Inc. Apparatus and method for matching image characteristics such as fingerprint minutiae
US5366898A (en) * 1992-03-27 1994-11-22 Dexsil Corporation Method for quantitative determination of total base or acid number of oil
US7043083B2 (en) 2001-03-28 2006-05-09 Nec Corporation Pattern-collating device, pattern-collating method and pattern-collating program
GB2487377B (en) * 2011-01-18 2018-02-14 Aptina Imaging Corp Matching interest points

Similar Documents

Publication Publication Date Title
KR100293623B1 (en) Figure location detecting system
US4542526A (en) Character recognition apparatus
WO2016127736A1 (en) Computing method for area of fingerprint overlapping area and electronic apparatus
JP2002298141A (en) Pattern collating device, pattern collating method thereof, and pattern collating program
CN111461113B (en) Large-angle license plate detection method based on deformed plane object detection network
JPS5924384A (en) Checking circuit of feature point
JPS59778A (en) Finger print collating device
JPH11167455A (en) Hand form recognition device and monochromatic object form recognition device
JPH11312243A (en) Facial region detector
JP2734245B2 (en) Image feature extraction method and apparatus
JPS585887A (en) Rotation angle detecting device for pattern
JP4304722B2 (en) Deformation estimation method, deformation estimation apparatus, and deformation estimation program
JP3106370B2 (en) Defect detection and type recognition of printed circuit boards using graph information
JP3116438B2 (en) Inspection apparatus and inspection method for printed wiring board
CN114239631B (en) Two-dimensional code identification method and device, storage medium and electronic equipment
JP2000036047A (en) Picture processor
JPS6073795A (en) Character recognition device
JP2921055B2 (en) Pattern matching device and pattern matching method
JPH0420228B2 (en)
Zhang et al. A novel arc segmentation approach for document image processing
JP2658153B2 (en) Character identification method
JPS622382A (en) Feature extracting devie for pattern
JPS634230B2 (en)
Ninomiya et al. Precise position detection of a mesh-backed printing screen based on the smallest enclosing circle detection
JPS6332673A (en) Picture recognizing device