WO2009081866A1 - パターン間特徴対応付け装置及びそれに用いるパターン間特徴対応付け方法並びにそのプログラム - Google Patents
パターン間特徴対応付け装置及びそれに用いるパターン間特徴対応付け方法並びにそのプログラム Download PDFInfo
- Publication number
- WO2009081866A1 WO2009081866A1 PCT/JP2008/073204 JP2008073204W WO2009081866A1 WO 2009081866 A1 WO2009081866 A1 WO 2009081866A1 JP 2008073204 W JP2008073204 W JP 2008073204W WO 2009081866 A1 WO2009081866 A1 WO 2009081866A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- feature point
- point group
- feature
- candidate
- arrangement
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1365—Matching; Classification
- G06V40/1371—Matching features related to minutiae or pores
Definitions
- the present invention relates to an inter-pattern feature associating apparatus, an inter-pattern feature associating method and a program thereof, and more particularly to a technique for associating feature points between patterns.
- This fingerprint collation device is a device for collating the identity of fingerprint patterns.
- the coordinates from the corresponding feature points as candidates to the most concentrated part by voting to the coordinate matching parameter space Determine the amount of alignment.
- the fingerprint matching device performs coordinate matching of feature points based on this coordinate matching amount, and after passing through a selection process for deleting candidate pairs that are greatly deviated, the corresponding feature points are determined.
- the coordinate matching process of the feature points performed to select candidate pairs treats the entire variation between the collated patterns as one rigid body motion. If an elastic strain fluctuation component that cannot be represented by the assumed geometric transformation is included, a part of the true corresponding feature point is deleted by this processing, and the corresponding feature point for authentication is not sufficiently extracted. .
- an object of the present invention is to solve the above-described problems and to provide an inter-pattern feature associating device capable of accurately determining a feature correspondence between patterns even when the pattern to be matched is distorted, and between the patterns used for the same.
- a feature association method and a program thereof are provided.
- the inter-pattern feature associating device includes: N feature points (N is an integer of 2 or more) that are close in position in the input pattern signal are used as a proximity feature point group, and the arrangement relationship between the proximity feature point group and the feature points of the proximity feature point group Generating means for generating at least one arrangement relation numerical value group representing as a feature point group arrangement; Detecting means for detecting corresponding corresponding feature point group arrangement candidates by comparing the arrangement relation numerical value of the first pattern and the arrangement relation numerical value of the second pattern generated by the generating means; Compare the feature relations between the feature points of each feature point group placement of the corresponding feature point group placement candidate and the feature points close to the feature point group placement, and add the corresponding neighboring feature points to the corresponding feature point group placement candidate And updating means for updating the corresponding feature point group placement candidate; Means for examining corresponding feature point candidates associated with the corresponding feature point group placement candidates updated by the updating means and determining corresponding feature points associated with the corresponding feature point group placement candidates; Means for inspecting the
- An inter-pattern feature matching method includes: N feature points (N is an integer of 2 or more) that are close in position in the input pattern signal are used as a proximity feature point group, and the arrangement relationship between the proximity feature point group and the feature points of the proximity feature point group; A first step of generating at least one arrangement relationship numerical value group representing the feature point group arrangement; A second step of detecting a corresponding feature point group arrangement candidate by comparing the arrangement relation numerical value of the first pattern and the arrangement relation numerical value of the second pattern generated in the first step; Compare the feature relations between the feature points of each feature point group placement of the corresponding feature point group placement candidate and the feature points close to the feature point group placement, and add the corresponding neighboring feature points to the corresponding feature point group placement candidate And a third step of updating the corresponding feature point group arrangement candidate, A fourth step of determining a corresponding feature point associated with the corresponding feature point group placement candidate by examining the corresponding feature point candidate associated with the corresponding feature point group placement candidate updated in the third step; And a fifth step of
- a program according to the present invention is a program to be executed by a central processing unit in an inter-pattern feature association apparatus that performs inter-pattern feature association using pattern feature point information, N feature points (N is an integer of 2 or more) that are close in position in the input pattern signal are used as a proximity feature point group, and the arrangement relationship between the proximity feature point group and the feature points of the proximity feature point group
- a first process for generating at least one arrangement relation numerical value group representing a feature point group arrangement A second process for detecting a corresponding feature point group arrangement candidate by comparing the arrangement relation numerical value of the first pattern and the arrangement relation numerical value of the second pattern generated in the first process; Compare the feature relations between the feature points of each feature point group placement of the corresponding feature point group placement candidate and the feature points close to the feature point group placement, and add the corresponding neighboring feature points to the corresponding feature point group placement candidate And a third process for updating the corresponding feature point group placement candidate;
- the present invention has the above-described configuration and operation, so that it is possible to accurately determine the feature correspondence between patterns even if the pattern to be verified is distorted.
- FIG. 1 is a block diagram showing a configuration example of an inter-pattern feature association apparatus according to the first embodiment of the present invention.
- the inter-pattern feature associating apparatus 1 includes a pattern proximity feature point group arrangement creating unit 11, a corresponding arrangement candidate detecting unit 12, a corresponding arrangement candidate updating unit 13, and a corresponding arrangement candidate corresponding feature point determining unit 14.
- FIG. 2 is a flowchart showing the operation of the inter-pattern feature associating method according to the first embodiment of the present invention. The operation of the inter-pattern feature associating method according to the first embodiment of the present invention will be described with reference to FIGS.
- the pattern proximity feature point group arrangement creation unit 11 of the inter-pattern feature association apparatus 1 receives the pattern signal and creates the arrangement of the proximity feature point group in the pattern signal (step S1 in FIG. 2).
- the corresponding arrangement candidate detection unit 12 inspects the coincidence of the adjacent feature point group arrangement created in the pattern signal between the input patterns in the pattern proximity feature point group arrangement generation unit 11 and detects the corresponding feature point group arrangement candidate (Step S2 in FIG. 2).
- the corresponding arrangement candidate update unit 13 examines the correspondence (feature point group) arrangement candidate detected by the corresponding arrangement candidate detection unit 12 to check the coincidence of the arrangement relation between each feature point group arrangement and its neighboring feature points. The adjacent feature points thus added are added, and the corresponding arrangement candidate is updated (step S3 in FIG. 2).
- the corresponding arrangement candidate corresponding feature point determination unit 14 examines each corresponding feature point candidate attached to the corresponding arrangement candidate for each corresponding arrangement candidate that has been updated by the corresponding arrangement candidate update unit 13, and the corresponding arrangement candidate. Corresponding feature points associated with are determined (step S4 in FIG. 2).
- Corresponding feature point determination unit 15 examines each corresponding arrangement candidate for which the corresponding corresponding feature point has been determined by corresponding arrangement candidate corresponding feature point determination unit 14 and derives the corresponding arrangement between patterns and the corresponding feature points associated therewith. (Step S5 in FIG. 2).
- FIGS. 4 (a) and 4 (b) are diagrams for explaining a method for generating the proximity feature point group arrangement in the first embodiment of the present invention.
- FIGS. 4 (a) and 4 (b) are diagrams of the present invention. It is a figure for demonstrating the detection method of the corresponding feature point group arrangement
- FIG. 5 (a), (b) is a corresponding feature point group arrangement in the 1st Embodiment of this invention. It is a figure for demonstrating the update method of a candidate.
- FIG. 6 (a) and 6 (b) are diagrams for explaining selection of associated corresponding feature point candidates in the first embodiment of the present invention
- FIG. 7 is a feature in the first embodiment of the present invention. It is a figure for demonstrating the arrangement
- the pattern proximity feature point group creation unit 11 receives a pattern signal and creates a layout of the proximity feature point group in the pattern signal.
- the arrangement of the feature point group includes a set of feature points detected in the pattern signal and an arrangement relation value group representing an arrangement relation between the feature points of the feature point set.
- the numerical value of the arrangement relationship between feature points includes numerical data that can determine the arrangement between feature points, such as the distance between feature points, the intersection angle between feature points, or the difference in feature point direction between feature points.
- the proximity feature point group is a set of feature points that are close in position within the pattern signal.
- the arrangement of all adjacent feature point groups is identified and managed by an identifier (ID: IDentifier). Furthermore, it is desirable that the number of feature points in the created proximity feature point group arrangement is 2 or more and a fixed integer N.
- FIGS. 3A and 3B schematically show partial enlarged views of the feature points of the fingerprint pattern and the proximity feature point group arrangement as an example of the pattern.
- FIG. 3A shows the spatial distribution of fingerprint feature points (P1 to P6).
- FIG. 3B shows an arrangement (D1 to D4) of adjacent feature point groups generated from these feature points and having three feature points.
- FIG. 7 schematically shows an example of the arrangement relation numerical value between feature points.
- the feature point direction can be detected for each feature point such as the distance L12 between the feature point P1 and the feature point P2, the intersection angle between the feature point vector P2P1 and the feature point vector P2P3, or a fingerprint pattern.
- the feature point direction v1 with respect to an arbitrary inter-feature point vector in the arrangement P1P2P3 of the feature points P1 can also be used as an arrangement relation numerical value between the feature points.
- Corresponding arrangement candidate detection unit 12 checks the coincidence of all feature point group arrangements created in the pattern signal between the input patterns, and detects the corresponding feature point group arrangement candidates. The coincidence check of the feature point group arrangement is performed based on whether or not the difference in the arrangement relation value between the feature points is within a predetermined allowable range. The corresponding arrangement candidate detection unit 12 registers the feature point group arrangement between patterns that have passed the consistency check as a corresponding arrangement candidate.
- the arrangement relation between the feature points used for checking the consistency of the feature point group arrangement may be measured using a general image coordinate system as a reference coordinate system.
- the measurement of the streamline coordinate system in the pattern proposed by the applicant of the present application using the reference coordinate system reduces the variation in the numerical value of the arrangement relation between feature points. .
- the streamline coordinate system is obtained by detecting a streamline in the pattern signal and setting a coordinate system (referred to as a “streamline coordinate system”) corresponding to the direction of the detected streamline.
- a coordinate system that is uniquely determined by the distribution of the flow direction of the pattern on the deformed striped pattern and that is defined by a group of curves consisting of a streamline curve and a normal curve orthogonal to each other is represented by a streamline coordinate system of the pattern (Also called “natural coordinate system”).
- the curve group is referred to as a “coordinate curve group”, and attributes such as the position and orientation of each point in the pattern in the streamline coordinate system are determined by the relationship between the point and each coordinate curve.
- FIG. 4A and 4 (b) are partial enlarged views of the respective feature points and proximity feature point group arrangements of the search pattern and the reference pattern to be compared in order to explain the detection of the corresponding arrangement candidate between patterns.
- FIG. 4A shows the proximity feature point group arrangements D1 to D4 of the search pattern
- FIG. 4B shows the proximity feature point group arrangements M1 to M3 of the reference pattern.
- (D1, M1), (D2, M2), and (D3, M3) each pass the consistency check and are detected as corresponding arrangement candidates.
- Corresponding placement candidate update unit 13 examines the correspondence between the feature points of each feature point group placement of the corresponding placement candidate and the feature point adjacent to the feature point group placement for the correspondence placement candidate, The proximity feature points that correspond according to the result are added to the corresponding feature point group arrangement as new corresponding feature point candidates. Further, such inspection / addition processing is sequentially performed on all adjacent feature points, and the corresponding arrangement candidates are continuously updated as much as possible.
- FIG. 5A and 5B schematically show arrangement candidates corresponding to a search pattern and a reference pattern and adjacent feature points of these arrangements.
- FIG. 5A shows the feature point group arrangement D1 of the search pattern and its neighboring feature points P4 to P6, and
- FIG. 5B shows the feature point group arrangement M1 of the reference pattern and its neighboring feature points Q4 to P6.
- Q5 is shown respectively.
- D1 and M1 are corresponding placement candidates that have passed the placement consistency check.
- the positional relation numerical value between the arrangement D1 and the feature point P4 is compared with the positional relation numerical value between the arrangement M1 and the feature point Q4, and there is a difference between the positional relation numerical values. If within the allowable range, P4 and Q4 are added to D1 and M1, respectively, and the corresponding arrangement (D1, M1) is replaced with a new corresponding arrangement (D1 ′, M1 ′).
- the intersection angle between the feature point direction v1 and the feature point vector, and the feature point may be compared.
- the corresponding arrangement candidate update unit 13 continuously inspects and updates the corresponding arrangement candidates, and records the corresponding arrangement that is no longer updated after all inspections are completed.
- the feature point pairs P1-Q1, P2-Q2, P3-Q3, P4-Q4, and P5-Q5 corresponding to the updated corresponding arrangement candidate (D1 ′′, M1 ′′) are associated with the corresponding arrangement candidate. Let it be a feature point candidate.
- Corresponding placement candidate corresponding feature point determination unit 14 examines each corresponding feature point candidate associated with the corresponding placement candidate with respect to the updated correspondence placement candidate, and determines a corresponding feature point of the corresponding placement candidate. Each corresponding feature point candidate is inspected based on a correspondence relationship between adjacent feature points of the corresponding feature point candidate.
- FIG. 6 (a) and 6 (b) show a corresponding arrangement in which the one-to-one property corresponding to the feature points as described above is not satisfied.
- the arrangement D1 in FIG. 6A and the arrangement M1 in FIG. 6B are corresponding arrangement candidates, and the correspondence relation between the proximity feature points (P4 and P7) of D1 and the proximity feature point (Q4) of M1 is determined.
- the selection is made based on the information of the corresponding arrangement (D1, M1), there is a high possibility that the selection is erroneous.
- the one-to-one property corresponding to the feature points may not be satisfied by the corresponding arrangement candidate update unit 13, and both P4-Q4 and P7-Q4 correspond to each other. Add it to the corresponding layout as a feature point.
- the degree of correspondence of the feature points in the proximity region centered on the feature point candidate for example, the corresponding feature It is better to leave a candidate with a high degree of correspondence based on the number of points.
- Corresponding feature point determination unit 15 evaluates each corresponding corresponding placement candidate for which the corresponding corresponding feature point has been determined, and derives the final corresponding placement and the corresponding corresponding feature point. For the evaluation of the corresponding arrangement candidate, for example, a method based on the number of corresponding feature points associated with the corresponding arrangement candidate is effective. In addition, once the corresponding arrangement is determined, the corresponding feature point determination unit 15 may output the corresponding feature point associated therewith as the final corresponding feature point.
- the influence of the distortion can be suppressed and the pattern identity can be verified.
- the present embodiment even if distortion occurs between the patterns to be verified, it is possible to extract the corresponding feature points between the patterns while suppressing the influence of the distortion.
- FIG. 9 is a block diagram showing a configuration example of the inter-pattern feature associating device according to the second embodiment of the present invention.
- the inter-pattern feature associating device 1 includes a CPU (central processing unit) 2 and a program storage unit 3 that stores a program executed by the CPU 2.
- the CPU 2 executes a program stored in the program storage unit 3, whereby the pattern proximity feature point group arrangement creating unit 11, the corresponding arrangement candidate detecting unit 12, and the corresponding arrangement candidate updating unit 13 as described above.
- the processing of the corresponding arrangement candidate corresponding feature point determination unit 14 and the corresponding feature point determination unit 15 is realized.
- the CPU 2 executes the program stored in the program storage unit 3 to achieve the same effect as that of the first embodiment of the present invention.
- FIG. 10 is a block diagram showing a configuration example of a pattern matching apparatus according to the third embodiment of the present invention.
- the pattern matching device 4 includes a search pattern feature storage unit 41, a reference pattern feature storage unit 42, a search pattern proximity feature point group location creation unit 11a, a reference pattern proximity feature point group location creation unit 11b,
- the corresponding arrangement candidate detection unit 12, the corresponding arrangement candidate update unit 13, the corresponding arrangement candidate corresponding feature point determination unit 14, the corresponding feature point determination unit 15, and the matching determination unit 43 are provided.
- the search pattern feature storage unit 41 holds both a pattern feature detected from a pattern to be searched and attribute data including position data of the pattern feature.
- the reference pattern feature storage unit 42 holds both pattern features detected from the reference pattern and attribute data including position data of the pattern features.
- the search pattern proximity feature point group creation unit 11a detects a combination of feature points that are close in position to the combination of pattern feature data held in the search pattern feature storage unit 41 as the search pattern side proximity feature point group placement.
- the reference pattern proximity feature point group arrangement creation unit 11b detects a combination of feature points that are close in position to the combination of pattern feature data held in the reference pattern feature storage unit 42 as a reference pattern side proximity feature point group arrangement. .
- the corresponding arrangement candidate detection unit 12 sequentially reads out the adjacent feature point arrangement data detected by the search pattern proximity feature point group arrangement creation unit 11a and the reference pattern proximity feature point group arrangement creation unit 11b, and sets the proximity feature point group arrangement. A relative distance between feature points, an intersection angle, a feature point direction, and the like are compared, and a feature point group arrangement corresponding to the omnibus is detected.
- the corresponding arrangement candidate update unit 13 inspects the coincidence of the relationship between the corresponding feature point group arrangement and the adjacent feature points of the feature point group arrangement, and determines the adjacent feature points according to the inspection result. To the corresponding feature point group arrangement, and the corresponding feature point group arrangement is updated.
- the corresponding arrangement candidate corresponding feature point determination unit 14 corresponds to each corresponding arrangement candidate that has been updated, based on the degree of correspondence of the adjacent feature points of the corresponding feature point candidate, with each corresponding feature point candidate associated with the corresponding arrangement candidate. Select feature point candidates and determine corresponding feature points.
- the corresponding feature point determination unit 15 evaluates each corresponding arrangement candidate for which the determination of the associated corresponding feature point has been completed, determines an appropriate corresponding arrangement, and outputs the corresponding feature point associated therewith.
- the collation determination unit 43 determines pattern matching from the corresponding feature point data output to the corresponding feature point determination unit 15. For example, a score may be derived from the number of corresponding feature points, and if the score exceeds a predetermined threshold, it may be determined that the pattern matching results match.
- the determination result in the matching determination unit 43 is, for example, permission to enter or leave a building or to enter or leave a building area. Used for etc.
- the search pattern proximity feature point group location creation unit 11a, the reference pattern proximity feature point group location creation unit 11b, the corresponding placement candidate detection unit 12, the corresponding placement candidate update unit 13, the corresponding placement candidate corresponding feature point determination unit 14, All or a part of the corresponding feature point determination unit 15 and the collation determination unit 43 may realize the processing / function by a program operating on a computer.
- the present invention can be applied to uses such as personal identification, identification confirmation, and suspect candidate search by collating finger / palm patterns.
- FIG. 1 It is a block diagram which shows the structural example of the pattern feature matching apparatus by the 1st Embodiment of this invention. It is a flowchart which shows operation
- (A), (b) is a figure for demonstrating the production
- (A), (b) is a figure for demonstrating the detection method of the corresponding feature point group arrangement
- (A), (b) is a figure for demonstrating the update method of the corresponding feature point group arrangement
- (A), (b) is a figure for demonstrating selection of the accompanying corresponding
- (A), (b) is a figure for demonstrating the comparison of the arrangement
- Pattern collation apparatus 11 Pattern proximity
- positioning preparation part 12 Corresponding arrangement
- positioning candidate detection part 13 Corresponding arrangement
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
Abstract
Description
入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する生成手段と、
前記生成手段で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する検出手段と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する更新手段と、
前記更新手段で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める手段と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める手段とを備えている。
入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する第1の工程と、
前記第1の工程で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する第2の工程と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する第3の工程と、
前記第3の工程で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める第4の工程と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める第5の工程とを含むことを特徴とする。
入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する第1の処理と、
前記第1の処理で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する第2の処理と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する第3の処理と、
前記第3の処理で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める第4の処理と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める第5の処理とを含むことを特徴とする。
2 CPU
3 プログラム記憶部
4 パターン照合装置
11 パターン近接特徴点群配置作成部
11a 探索パターン近接特徴点群配置作成部
11b 参照パターン近接特徴点群配置作成部
12 対応配置候補検出部
13 対応配置候補更新部
14 対応配置候補対応特徴点決定部
15 対応特徴点決定部
41 探索パターン特徴記憶部
42 参照パターン特徴記憶部
43 照合判定部
Claims (17)
- 入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する生成手段と、
前記生成手段で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する検出手段と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する更新手段と、
前記更新手段で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める手段と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める手段とを有することを特徴とするパターン間特徴対応付け装置。 - 前記パターン信号内の流線を検出しかつ検出した流線の方向に対応した流線座標系を設定する手段を含み、
前記特徴点間の配置関係数値は、前記流線座標系を基準座標系として求めることを特徴とする請求項1記載のパターン間特徴対応付け装置。 - 前記特徴点間の配置関係数値は、
一方の特徴点から見た他方の特徴点の基準座標系での距離、
一方の特徴点から見た他方の二つの特徴点の基準座標系での方角のずれ、
一方の特徴点から見た他方の特徴点の基準座標系での特徴点方向のずれ、
のうちの少なくとも1つを含むことを特徴とする請求項1記載のパターン間特徴対応付け装置。 - 前記対応特徴点群配置候補の各特徴点群配置の第1の特徴点と当該対応特徴点群配置に近接する第2の近接特徴点との距離、当該第1の特徴点から見た当該第2の近接特徴点と他方の当該対応特徴点群配置の第3の特徴点との方角のずれをそれぞれ比較し、対応する近接特徴点を前記対応特徴点群配置候補に追加し、前記対応特徴点群配置候補を更新することを特徴とする請求項1記載のパターン間特徴対応付け装置。
- 前記対応特徴点群配置候補の各特徴点群配置の第1の特徴点と当該特徴点群配置に近接する第2の近接特徴点との距離、当該第1の特徴点に対する当該第2の近接特徴点の特徴点方向のずれをそれぞれ比較し、対応する近接特徴点を前記対応特徴点群配置候補に追加し、前記対応特徴点群配置候補を更新することを特徴とする請求項1記載のパターン間特徴対応付け装置。
- 前記更新手段で更新された対応特徴点群配置候補の付随する対応特徴点候補に対する検査を、1対1の性質が満たされていない対応特徴点候補に対して行うことを特徴とする請求項1記載のパターン間特徴対応付け装置。
- 前記更新手段で更新された対応特徴点群配置候補の付随する対応特徴点候補に対する検査が、対応特徴点候補の近接する特徴点の対応関係に基づくことを特徴とする請求項1記載のパターン間特徴対応付け装置。
- 前記対応特徴点群配置候補に対する検査が、前記対応特徴点群配置候補の付随する対応特徴点の個数に基づくことを特徴とする請求項1記載のパターン間特徴対応付け装置。
- 入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する第1の工程と、
前記第1の工程で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する第2の工程と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する第3の工程と、
前記第3の工程で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める第4の工程と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める第5の工程とを含むことを特徴とするパターン間特徴対応付け方法。 - 前記パターン信号内の流線を検出しかつ検出した流線の方向に対応した流線座標系を設定する工程を含み、
前記特徴点間の配置関係数値は、前記流線座標系を基準座標系として求めることを特徴とする請求項9記載のパターン間特徴対応付け方法。 - 前記特徴点間の配置関係数値は、
一方の特徴点から見た他方の特徴点の基準座標系での距離、
一方の特徴点から見た他方の二つの特徴点の基準座標系での方角のずれ、
一方の特徴点から見た他方の特徴点の基準座標系での特徴点方向のずれ、
のうちの少なくとも1つを含むことを特徴とする請求項9記載のパターン間特徴対応付け方法。 - 前記対応特徴点群配置候補の各特徴点群配置の第1の特徴点と当該対応特徴点群配置に近接する第2の近接特徴点との距離、当該第1の特徴点から見た当該第2の近接特徴点と他方の当該対応特徴点群配置の第3の特徴点との方角のずれをそれぞれ比較し、対応する近接特徴点を前記対応特徴点群配置候補に追加し、前記対応特徴点群配置候補を更新することを特徴とする請求項9記載のパターン間特徴対応付け方法。
- 前記対応特徴点群配置候補の各特徴点群配置の第1の特徴点と当該特徴点群配置に近接する第2の近接特徴点との距離、当該第1の特徴点に対する当該第2の近接特徴点の特徴点方向のずれをそれぞれ比較し、対応する近接特徴点を前記対応特徴点群配置候補に追加し、前記対応特徴点群配置候補を更新することを特徴とする請求項9記載のパターン間特徴対応付け方法。
- 前記第3の工程で更新された対応特徴点群配置候補の付随する対応特徴点候補に対する検査を、1対1の性質が満たされていない対応特徴点候補に対して行うことを特徴とする請求項9記載のパターン間特徴対応付け方法。
- 前記第3の工程で更新された対応特徴点群配置候補の付随する対応特徴点候補に対する検査が、対応特徴点候補の近接する特徴点の対応関係に基づくことを特徴とする請求項9記載のパターン間特徴対応付け方法。
- 前記対応特徴点群配置候補に対する検査が、前記対応特徴点群配置候補の付随する対応特徴点の個数に基づくことを特徴とする請求項9記載のパターン間特徴対応付け方法。
- パターンの特徴点情報を用いてパターン間特徴の対応付けを行うパターン間特徴対応付け装置内の中央処理装置に実行させるプログラムであって、
入力されたパターン信号内の位置的に近接するN個(Nは2以上の整数)の特徴点を近接特徴点群としかつ前記近接特徴点群と当該近接特徴点群の特徴点間の配置関係を表わす配置関係数値群とを特徴点群配置として少なくとも1つ生成する第1の処理と、
前記第1の処理で生成された第1パターンの配置関係数値と第2パターンの配置関係数値とを比較して対応する対応特徴点群配置候補を検出する第2の処理と、
前記対応特徴点群配置候補の各特徴点群配置の特徴点と当該特徴点群配置に近接する特徴点との配置関係数値を比較して対応する近接特徴点を対応特徴点群配置候補に追加しかつ前記対応特徴点群配置候補を更新する第3の処理と、
前記第3の処理で更新された対応特徴点群配置候補の付随する対応特徴点候補を検査して前記対応特徴点群配置候補の付随する対応特徴点を定める第4の処理と、
前記対応特徴点群配置候補を検査してパターン間の対応特徴点を定める第5の処理とを含むことを特徴とするプログラム。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2009547079A JP5304655B2 (ja) | 2007-12-26 | 2008-12-19 | パターン間特徴対応付け装置及びそれに用いるパターン間特徴対応付け方法並びにそのプログラム |
EP08863448.0A EP2237226A4 (en) | 2007-12-26 | 2008-12-19 | DEVICE FOR MATCHING CHARACTERISTICS BETWEEN PATTERNS, METHOD FOR MATCHING CHARACTERISTICS BETWEEN PATTERNS USED FOR THE SAME, AND PROGRAM RELATING THERETO |
CN2008801229010A CN101911115B (zh) | 2007-12-26 | 2008-12-19 | 图案间特征对应设备、其图案间特征对应方法 |
US12/809,287 US8406486B2 (en) | 2007-12-26 | 2008-12-19 | Inter-pattern feature corresponding device, inter-pattern feature corresponding method used for the same, and program therefor |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2007333390 | 2007-12-26 | ||
JP2007-333390 | 2007-12-26 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2009081866A1 true WO2009081866A1 (ja) | 2009-07-02 |
Family
ID=40801167
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2008/073204 WO2009081866A1 (ja) | 2007-12-26 | 2008-12-19 | パターン間特徴対応付け装置及びそれに用いるパターン間特徴対応付け方法並びにそのプログラム |
Country Status (5)
Country | Link |
---|---|
US (1) | US8406486B2 (ja) |
EP (1) | EP2237226A4 (ja) |
JP (1) | JP5304655B2 (ja) |
CN (1) | CN101911115B (ja) |
WO (1) | WO2009081866A1 (ja) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2354999A1 (en) * | 2010-02-01 | 2011-08-10 | Daon Holdings Limited | Method and system for biometric authentication |
EP2354998A1 (en) * | 2010-02-01 | 2011-08-10 | Daon Holdings Limited | Method and system of accounting for positional variability of biometric features |
US8041956B1 (en) | 2010-08-16 | 2011-10-18 | Daon Holdings Limited | Method and system for biometric authentication |
JP2013137608A (ja) * | 2011-12-28 | 2013-07-11 | Azbil Corp | 評価装置および評価方法 |
JP2013196418A (ja) * | 2012-03-21 | 2013-09-30 | Dainippon Printing Co Ltd | 個体識別装置、個体識別方法、プログラム |
WO2014061221A1 (ja) * | 2012-10-18 | 2014-04-24 | 日本電気株式会社 | 画像部分領域抽出装置、画像部分領域抽出方法および画像部分領域抽出用プログラム |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2495687B1 (en) | 2011-03-02 | 2018-12-05 | Precise Biometrics AB | Method of matching, biometric matching apparatus, and computer program |
JP2012256272A (ja) * | 2011-06-10 | 2012-12-27 | Seiko Epson Corp | 生体識別装置、及び、生体識別方法 |
KR101901588B1 (ko) * | 2012-01-02 | 2018-09-28 | 삼성전자주식회사 | 물체 인식 방법과, 물체 인식을 위한 기술자 생성 방법, 물체 인식을 위한 기술자 생성 장치 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS59778A (ja) | 1982-06-28 | 1984-01-05 | Nec Corp | 指紋照合装置 |
JPH06103366A (ja) * | 1992-09-18 | 1994-04-15 | Matsumura Electron:Kk | 指紋照合装置及び指紋照合方法 |
JPH10240932A (ja) * | 1997-02-27 | 1998-09-11 | Nec Software Ltd | 縞パターン照合装置 |
JPH11110542A (ja) | 1997-09-30 | 1999-04-23 | Toshiba Corp | パターン抽出方法および装置、そのプログラムを記録した媒体 |
JPH11195119A (ja) | 1997-10-20 | 1999-07-21 | Fujitsu Ltd | 指紋登録装置、指紋照合装置及び指紋照合方法 |
JP2007333390A (ja) | 2006-06-12 | 2007-12-27 | Yokogawa Electric Corp | 光信号モニタ |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5613014A (en) * | 1994-10-12 | 1997-03-18 | Martin Marietta Corp. | Fingerprint matching system |
US6041133A (en) * | 1996-12-13 | 2000-03-21 | International Business Machines Corporation | Method and apparatus for fingerprint matching using transformation parameter clustering based on local feature correspondences |
US6330347B1 (en) * | 1997-02-19 | 2001-12-11 | Stmicroelectronics S.R.L. | Method and device for identifying fingerprints using an analog flash memory |
JP3725998B2 (ja) * | 1999-10-14 | 2005-12-14 | 富士通株式会社 | 指紋照合装置及び照合方法 |
JP3241028B2 (ja) * | 1999-12-22 | 2001-12-25 | 日本電気株式会社 | 画像撮像装置及びその画像撮像方法 |
US6778687B2 (en) * | 2001-04-24 | 2004-08-17 | Lockheed Martin Corporation | Fingerprint matching system with ARG-based prescreener |
US7046829B2 (en) * | 2001-05-30 | 2006-05-16 | International Business Machines Corporation | Fingerprint verification |
JP3719435B2 (ja) * | 2002-12-27 | 2005-11-24 | セイコーエプソン株式会社 | 指紋照合方法及び指紋照合装置 |
EP1524620B1 (en) * | 2003-10-17 | 2007-08-15 | Berner Fachhochschule, Technik und Informatik (TI) | Method to conduct fingerprint verification and a fingerprint verification system |
KR100837002B1 (ko) * | 2004-01-15 | 2008-06-10 | 닛본 덴끼 가부시끼가이샤 | 패턴 식별 시스템, 패턴 식별 방법, 및 패턴 식별 프로그램을 기록한 기록매체 |
JP4886371B2 (ja) * | 2006-06-07 | 2012-02-29 | 株式会社日立情報制御ソリューションズ | 生体認証方法及びシステム |
CN100447806C (zh) * | 2006-06-14 | 2008-12-31 | 北京握奇数据系统有限公司 | 两阶段混合指纹特征匹配方法、匹配装置及其应用 |
-
2008
- 2008-12-19 WO PCT/JP2008/073204 patent/WO2009081866A1/ja active Application Filing
- 2008-12-19 US US12/809,287 patent/US8406486B2/en active Active
- 2008-12-19 JP JP2009547079A patent/JP5304655B2/ja active Active
- 2008-12-19 CN CN2008801229010A patent/CN101911115B/zh active Active
- 2008-12-19 EP EP08863448.0A patent/EP2237226A4/en not_active Ceased
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS59778A (ja) | 1982-06-28 | 1984-01-05 | Nec Corp | 指紋照合装置 |
JPH06103366A (ja) * | 1992-09-18 | 1994-04-15 | Matsumura Electron:Kk | 指紋照合装置及び指紋照合方法 |
JPH10240932A (ja) * | 1997-02-27 | 1998-09-11 | Nec Software Ltd | 縞パターン照合装置 |
JPH11110542A (ja) | 1997-09-30 | 1999-04-23 | Toshiba Corp | パターン抽出方法および装置、そのプログラムを記録した媒体 |
JPH11195119A (ja) | 1997-10-20 | 1999-07-21 | Fujitsu Ltd | 指紋登録装置、指紋照合装置及び指紋照合方法 |
JP2007333390A (ja) | 2006-06-12 | 2007-12-27 | Yokogawa Electric Corp | 光信号モニタ |
Non-Patent Citations (1)
Title |
---|
See also references of EP2237226A4 |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2354999A1 (en) * | 2010-02-01 | 2011-08-10 | Daon Holdings Limited | Method and system for biometric authentication |
EP2354998A1 (en) * | 2010-02-01 | 2011-08-10 | Daon Holdings Limited | Method and system of accounting for positional variability of biometric features |
US8520903B2 (en) | 2010-02-01 | 2013-08-27 | Daon Holdings Limited | Method and system of accounting for positional variability of biometric features |
US8041956B1 (en) | 2010-08-16 | 2011-10-18 | Daon Holdings Limited | Method and system for biometric authentication |
US8977861B2 (en) | 2010-08-16 | 2015-03-10 | Daon Holdings Limited | Method and system for biometric authentication |
JP2013137608A (ja) * | 2011-12-28 | 2013-07-11 | Azbil Corp | 評価装置および評価方法 |
JP2013196418A (ja) * | 2012-03-21 | 2013-09-30 | Dainippon Printing Co Ltd | 個体識別装置、個体識別方法、プログラム |
WO2014061221A1 (ja) * | 2012-10-18 | 2014-04-24 | 日本電気株式会社 | 画像部分領域抽出装置、画像部分領域抽出方法および画像部分領域抽出用プログラム |
Also Published As
Publication number | Publication date |
---|---|
JP5304655B2 (ja) | 2013-10-02 |
CN101911115B (zh) | 2012-08-08 |
US20110200261A1 (en) | 2011-08-18 |
JPWO2009081866A1 (ja) | 2011-05-06 |
US8406486B2 (en) | 2013-03-26 |
EP2237226A4 (en) | 2015-09-30 |
EP2237226A1 (en) | 2010-10-06 |
CN101911115A (zh) | 2010-12-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5304655B2 (ja) | パターン間特徴対応付け装置及びそれに用いるパターン間特徴対応付け方法並びにそのプログラム | |
JP5020395B2 (ja) | 画像処理装置および画像処理方法 | |
Chikkerur et al. | K-plet and coupled BFS: a graph based fingerprint representation and matching algorithm | |
US7403634B2 (en) | Object tracking apparatus and method | |
US8520981B2 (en) | Document retrieval of feature point groups using a geometrical transformation | |
JP4877392B2 (ja) | 特徴属性計算装置、特徴量抽出装置とパターン照合装置と方法及びプログラム | |
Fan et al. | Singular points detection based on zero-pole model in fingerprint images | |
JP2008257469A (ja) | 画像処理装置及び画像検索方法 | |
JP2007265179A (ja) | レイアウト検証方法、レイアウト検証装置 | |
JP5958460B2 (ja) | 特徴点照合装置、特徴点照合方法、および特徴点照合プログラム | |
JP2002288667A (ja) | パターン照合装置とそのパターン照合方法、及びパターン照合プログラム | |
US20080019568A1 (en) | Object tracking apparatus and method | |
JP4032241B2 (ja) | 指紋照合装置及びその方法 | |
JP3884468B2 (ja) | 高速画像探索方法 | |
JP5566257B2 (ja) | データ生成方法および画像検査方法 | |
JP3819828B2 (ja) | 微細パターン測定方法 | |
US8219948B2 (en) | Layout verification device, layout verification program, and layout verification method of layout pattern of semiconductor device | |
JP3884462B2 (ja) | 高速画像探索方法 | |
JP3460824B2 (ja) | 形状認識方法及び装置 | |
JP4304722B2 (ja) | 変形推定方法、変形推定装置及び変形推定プログラム | |
JP4525526B2 (ja) | パターンマッチング方法及び装置 | |
JP2007102422A (ja) | パターン照合装置 | |
US20190205613A1 (en) | Distorted fingerprint matching using pose and minutia grouping | |
JP2012118732A (ja) | 3次元形状認識装置及び方法 | |
Gao et al. | Election Based Pose Estimation of Moving Objects |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200880122901.0 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 08863448 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2009547079 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 12809287 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2008863448 Country of ref document: EP |