JP4024960B2

JP4024960B2 - Image collation device and image collation method

Info

Publication number: JP4024960B2
Application number: JP16090099A
Authority: JP
Inventors: 和章横田
Original assignee: Toshiba Corp
Current assignee: Toshiba Corp
Priority date: 1999-06-08
Filing date: 1999-06-08
Publication date: 2007-12-19
Anticipated expiration: 2019-06-08
Also published as: JP2000348139A

Description

【０００１】
【発明の属する技術分野】
この発明は、たとえば筆跡照合などに適用して好適な画像照合装置および画像照合方法に係り、特に、画像照合の処理時間を大幅に短縮し、かつ、被照合画像の補正時に発生する直線パターンの誤消去を防止する画像照合装置および画像照合方法に関する。
【０００２】
【従来の技術】
近年のコンピュータ技術の向上に伴ない、高度な画像処理が可能となり、たとえば２枚の画像を比較して一致または不一致を判定することにより指紋や筆跡などを照合する画像照合システムが種々開発されている。そして、この画像を照合する代表的な方式の１つに、２枚の画像それぞれから特徴を抽出して比較する方式がある。この方式は、特徴の抽出精度の問題があるため、極めて類似した画像のわずかな差異部分を検出するのには向かないが、一般的に高速である。
【０００３】
また、この他に、２枚の画像中の各ピクセルを直接比較する方式がある。この方式では、画像のずれ、伸縮、回転などを補正して比較しなければならないために低速だが、画像のわずかな差異部分を検出するのにも適している。たとえば特開平１１−３４３１号などはこの方式で画像の比較を行なうものである。
【０００４】
【発明が解決しようとする課題】
しかしながら、この２枚の画像中の各ピクセルを直接比較する方式は、基準画像を矩形の小領域で複数に分割し、同じ特長をもつ小領域を被照合画像から検索するものであり、基準画像と被照合画像との各ピクセルの対応関係を調査するのに非常に大きな時間を費やしてしまうといった欠点があった。
【０００５】
また、基準画像を矩形の小領域で複数に分割するために、基準画像中の直線パターンを誤って消去してしまうこともあった。
【０００６】
この発明はこのような実情を考慮してなされたものであり、画像照合の処理時間を大幅に短縮し、かつ、被照合画像の補正時に発生する直線パターンの誤消去を防止する画像照合装置および画像照合方法を提供することを目的とする。
【０００７】
【課題を解決するための手段】
第１の発明は、前述した目的を達成するために、比較する２つの画像を小領域で複数に分割する補正処理前に、この２つの画像から罫線を検出し、その検出した罫線の２つの画像間における対応関係を調査し、その調査した対応関係から２つの画像を変形するといった補正処理を行なうようにしたものである。
【０００８】
この第１の発明においては、２つの画像の各ピクセルの対応関係を調査するといった細かな補正処理の前処理として、２つの画像の罫線の対応関係を調査するといった大まかな補正処理を実行するため、画像照合の処理時間を大幅に短縮することを可能とする。
【０００９】
また、第２の発明は、前述した目的を達成するために、比較する２つの画像を小領域で複数に分割する際、この小領域の配列が六方晶状の配列になるようにしたものである。
【００１０】
この発明においては、従来のように矩形の小領域で画像を分割する場合と異なり、隣り合う小領域間の境界線が連続する直線を形成することがないため、被照合画像の補正時に発生する直線パターンの誤消去を防止することを可能とする。
【００１１】
【発明の実施の形態】
以下、図面を参照してこの発明の実施形態を説明する。図１は、この発明の実施形態に係る画像照合装置の機器構成を示す図である。
【００１２】
図１に示すように、この画像照合装置は、ＣＰＵ１、システムメモリ２およびＩ／Ｏコントローラ３を備えている。ＣＰＵ１は、この画像照合装置全体の制御を司るものであり、システムメモリ２に格納された各種プログラムを実行制御する。システムメモリ２は、ＣＰＵ１によって実行制御される各種プログラムおよび各種処理データを格納する記憶装置である。Ｉ／Ｏコントローラ３は、基準画像や被照合画像を入力し、あるいはこれらの差分画像を出力するためのデータ入出力を制御する装置である。そして、この画像照合装置の画像照合処理は、システムメモリ２に格納されＣＰＵ１によって実行制御されるプログラム（画像照合システム）によって実現されている。
【００１３】
（第１実施形態）
まず、第１実施形態を説明する。
【００１４】
図２は、第１実施形態の画像照合装置の処理の流れを示したものである。この画像照合装置は、まず、イメージスキャナ等から読み取った入力画像から罫線を抽出する（ステップＡ１）。そして、この入力画像から抽出した罫線と基準画像から抽出された罫線とを比較し、入力画像と基準画像とのピクセルの大まかな対応を調査する（ステップＡ２）。
【００１５】
次に、画像照合装置は、入力画像を小領域に分割し、調査した対応により算出される基準画像上の位置付近から、もっとも近いパターンを持つ領域を検索して写像する（ステップＡ３）。そして、この写像した画像を小領域に分割し、各小領域を移動させることで基準画像と最も一致する画像を得（ステップＡ４）、最後に得た画像と基準画像との差分画像を出力する（ステップＡ５）。
【００１６】
以下、この各ステップの処理を詳説する。
【００１７】
［罫線の抽出（ステップＡ１）］
罫線の抽出にはいくつかの手法が既に確立しており、たとえば特開平１０−２４０９５９号や特開平６−６０２２２号等に述べられている方法がある。この画像照合装置では、これらの方法を用いて画像から罫線を抽出する。
【００１８】
［罫線の対応関係の調査（ステップＡ２）］
この画像照合装置は、基準画像から抽出した罫線と入力画像から抽出した罫線とを比較し、それらの対応を調査する。これには、罫線の相対距離を評価関数とした幅優先検索法等を使用する。
【００１９】
［入力画像の写像の作成（ステップＡ３）］
いま、入力画像上の４つの罫線および基準画像上の４つの罫線が図３に示すように対応するものとする。ここでは、基準画像のスキューは取り除いてあり、４つの罫線は用紙の方向に対して平行もしくは直角であるものとする。この時、図４に示す各交点の座標は次の（１）式〜（６）式のように計算できる。
【００２０】
【数１】

【００２１】
すると、基準画像上の座標（ｘ，ｙ）にあるピクセルは、次の（７）式で示される入力画像上の座標（Ｘ，Ｙ）と対応する。
【００２２】
【数２】

【００２３】
ここで、ｄ1 …ｄ4およびＸm，Ｙmは、次の（８）式〜（１３）式により計算される。
【００２４】
【数３】

【００２５】
そして、この画像照合装置は、これらの式を用いて、入力画像の各ピクセルを基準画像上のピクセルに写像する。この操作は、入力画像を拡大、縮小または回転して基準画像に一致するよう変形することを意味する。
【００２６】
ところで、前述の特開平１１−３４３１号では、基準画像上の小領域のパターンを入力画像から検索することでこの処理を行なっている。このような小領域のパターンを検索する作業には、２次元的な積算演算処理が必要である。一方、この画像照合装置が行なう罫線の検出処理は、１次元的な積算演算処理で実現可能である。このため、より高い処理速度が得られることになる。
【００２７】
［写像画像の微修正（ステップＡ４）］
前述の手順で写像した画像は、特に距離の離れた罫線により位置合わせを行なう場合、細部にわずかなずれが生じる。そこで、この画像照合装置では、写像後の画像を小領域に区切り、図５のように上下左右数ピクセルの範囲で小領域を動かして最も基準画像と一致する画像を得る。
【００２８】
従来、この処理を行なう場合、画像を矩形小領域に分割していた。しかしながら、このような分割を行なうと、直線が多く現れる画像において障害が発生する場合があった。
【００２９】
図６に、この障害が起こる画像の例を示す。図６（ａ）は従来の小領域の形状を示しており、図６（ｂ）の画像では、写像後の画像の小領域の境界部分に線状のパターン（中央横方向の太線）が存在する。ここで、この矩形の小領域を上下左右１ピクセルのみ移動可能とすると、図６（ｄ）の基準画像と最も一致するように各小領域を動すことにより、図６（ｃ）の画像が得られる。したがって、直線のように入力画像にしばしば現れるパターンがこの処理により消去されてしまう。このため、入力画像と基準画像の明確な差異である線が検出されないことになる。
【００３０】
そこで、この画像照合装置では、この障害を回避するために、小領域の形状として図７に示すような六方晶状の形状を採用する。六角形は平面を敷き詰めることができる最大の多角形であり、敷き詰めた際の境界線には連続的な直線パターンはない。したがって、直線的なパターンが完全に消去されることはない。
【００３１】
［基準画像との比較（ステップＡ５）］
最後に、この画像照合装置は、前述の手順で得られた画像と基準画像とを比較する。この際、スキャナ等から読み取った画像にはわずかなピクセルのぶれがあるために、図８のように、共通ピクセルの周囲にある差異ピクセルは差異ピクセルとみなさないようにする。この処理を行なうことで、最終的にはっきりとした差分画像を得ることができる。
【００３２】
（第２実施形態）
次に、第２実施形態を説明する。図９は、第２実施形態の画像照合装置の処理の流れを示したものである。この第２実施形態の画像照合装置は、２枚の入力画像と１枚の基準画像との計３枚の画像により処理を行なう。より具体的には、たとえば基準画像として白紙帳票画像を用意し、一方、入力画像として記入済みの帳票画像とさらに修正を加えた帳票画像とを用意し、この修正部分のみを検知する処理などを行なう。
【００３３】
この第２実施形態では、画像照合装置は、２つの入力画像のそれぞれについて拡大、縮小、回転までの処理を行ない（ステップＢ１〜ステップＢ６）、その後２つの画像が最も一致するよう微修正して（ステップＢ７）、最後に差分画像を出力する（ステップＢ８）。このため、スキューを含む２枚の画像を基準画像に一致するよう位置合わせを行なった上で比較を行なうことができる。
【００３４】
【発明の効果】
以上詳述したように、この発明によれば、前処理として罫線を検知し、対応を調べ、調べた画像から拡大・縮小・回転を行なうことにより、画像照合の処理時間を大幅に短縮することが可能となる。
【００３５】
また、分割後の小領域を六方晶状に配列するようにしたことにより、入力画像に頻繁に現れる直線状のパターンの誤消去を防止することが可能となる。
【図面の簡単な説明】
【図１】この発明の実施形態に係る画像照合装置の機器構成を示す図。
【図２】同第１実施形態の画像照合装置の処理の流れを示す図。
【図３】同第１実施形態の画像照合装置が処理する入力画像上の４つの罫線および基準画像上の４つの罫線の対応関係を示す図。
【図４】同第１実施形態の画像照合装置が処理する入力画像上の４つの罫線および基準画像上の４つの罫線の各交点の座標を示す図。
【図５】同第１実施形態の画像照合装置の写像画像の微修正を説明するための図。
【図６】同第１実施形態の画像照合装置の写像画像の微修正における問題を説明するための図。
【図７】同第１実施形態の画像照合装置の写像画像の微修正における問題を説明するための図。
【図８】同第１実施形態の画像照合装置の共通ピクセルの周囲にある差異ピクセルを無視した比較を説明するための図。
【図９】同第２実施形態の画像照合装置の処理の流れを示す図。
【符号の説明】
１…ＣＰＵ
２…システムメモリ
３…Ｉ／Ｏコントローラ
１１…画像照合システム[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image collation apparatus and an image collation method suitable for application to, for example, handwriting collation, and in particular, significantly shortens the processing time of image collation and generates a linear pattern generated when correcting the collated image. The present invention relates to an image collation apparatus and an image collation method for preventing erroneous erasure.
[0002]
[Prior art]
With recent improvements in computer technology, advanced image processing has become possible. For example, various image collation systems have been developed that collate fingerprints and handwriting by comparing two images to determine whether they match or do not match. Yes. As one of the typical methods for collating the images, there is a method of extracting and comparing features from each of the two images. This method is not suitable for detecting a slight difference portion of very similar images because of the problem of feature extraction accuracy, but is generally fast.
[0003]
In addition, there is a method in which each pixel in two images is directly compared. This method is slow because it must be corrected and compared for image shift, expansion and contraction, rotation, etc., but is also suitable for detecting slight differences in the image. For example, Japanese Patent Laid-Open No. 11-3431 and the like compare images by this method.
[0004]
[Problems to be solved by the invention]
However, the method of directly comparing each pixel in the two images is to divide the reference image into a plurality of rectangular small areas and search for small areas having the same features from the image to be matched. In other words, it takes a very long time to investigate the correspondence between each pixel and the image to be verified.
[0005]
Further, in order to divide the reference image into a plurality of rectangular small areas, the straight line pattern in the reference image may be mistakenly deleted.
[0006]
The present invention has been made in consideration of such a situation, and an image matching apparatus that significantly shortens the processing time of image matching and prevents erroneous erasure of a linear pattern that occurs during correction of the image to be verified. An object is to provide an image matching method.
[0007]
[Means for Solving the Problems]
In order to achieve the above-mentioned object, the first invention detects a ruled line from these two images before the correction processing for dividing the two images to be compared into a plurality of small areas, and detects the two ruled lines detected. Correspondence between images is investigated, and correction processing is performed such that two images are deformed from the investigated correspondence.
[0008]
In the first invention, in order to execute rough correction processing such as investigating the correspondence between ruled lines of two images as pre-processing for fine correction processing such as investigating the correspondence between pixels of two images. Thus, it is possible to greatly reduce the processing time of image collation.
[0009]
In order to achieve the above-mentioned object, the second invention is such that when two images to be compared are divided into a plurality of small areas, the arrangement of the small areas is a hexagonal arrangement. is there.
[0010]
In the present invention, unlike the conventional case in which an image is divided into rectangular small areas, a straight line in which the boundary line between adjacent small areas does not form a continuous line is generated, and this occurs at the time of correcting the image to be verified. It is possible to prevent erroneous erasure of the linear pattern.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a device configuration of an image collating apparatus according to an embodiment of the present invention.
[0012]
As shown in FIG. 1, the image collating apparatus includes a CPU 1, a system memory 2, and an I / O controller 3. The CPU 1 controls the entire image collating apparatus, and controls the execution of various programs stored in the system memory 2. The system memory 2 is a storage device that stores various programs executed and controlled by the CPU 1 and various processing data. The I / O controller 3 is a device that controls input / output of data for inputting a reference image and an image to be collated or outputting a difference image thereof. The image collation processing of the image collation apparatus is realized by a program (image collation system) stored in the system memory 2 and executed and controlled by the CPU 1.
[0013]
(First embodiment)
First, the first embodiment will be described.
[0014]
FIG. 2 shows the flow of processing of the image collating apparatus of the first embodiment. The image collating apparatus first extracts ruled lines from an input image read from an image scanner or the like (step A1). Then, the ruled line extracted from the input image is compared with the ruled line extracted from the reference image, and a rough correspondence of pixels between the input image and the reference image is investigated (step A2).
[0015]
Next, the image matching device divides the input image into small regions, searches for a region having the closest pattern from the vicinity of the position on the reference image calculated by the investigated correspondence, and maps the region (step A3). Then, this mapped image is divided into small areas, and each small area is moved to obtain an image that most closely matches the reference image (step A4), and a difference image between the last obtained image and the reference image is output. (Step A5).
[0016]
Hereinafter, the processing of each step will be described in detail.
[0017]
[Rule Extraction (Step A1)]
Several methods have already been established for the extraction of ruled lines. For example, there are methods described in JP-A-10-240959 and JP-A-6-60222. In this image collation apparatus, ruled lines are extracted from an image using these methods.
[0018]
[Investigation of ruled line correspondence (step A2)]
This image collating device compares the ruled lines extracted from the reference image with the ruled lines extracted from the input image, and investigates their correspondence. For this, a width-first search method or the like using the relative distance between ruled lines as an evaluation function is used.
[0019]
[Create Input Image Mapping (Step A3)]
Now, it is assumed that the four ruled lines on the input image and the four ruled lines on the reference image correspond as shown in FIG. Here, it is assumed that the skew of the reference image is removed and the four ruled lines are parallel or perpendicular to the paper direction. At this time, the coordinates of each intersection shown in FIG. 4 can be calculated as the following equations (1) to (6).
[0020]
[Expression 1]

[0021]
Then, the pixel at the coordinates (x, y) on the reference image corresponds to the coordinates (X, Y) on the input image represented by the following equation (7).
[0022]
[Expression 2]

[0023]
Here, d1... D4 and Xm and Ym are calculated by the following equations (8) to (13).
[0024]
[Equation 3]

[0025]
And this image collation apparatus maps each pixel of an input image to the pixel on a reference | standard image using these formulas. This operation means that the input image is enlarged, reduced, or rotated so as to be matched with the reference image.
[0026]
By the way, in the above-mentioned Japanese Patent Application Laid-Open No. 11-3431, this processing is performed by retrieving a pattern of a small area on a reference image from an input image. The operation of searching for such a small area pattern requires two-dimensional integration calculation processing. On the other hand, the ruled line detection process performed by the image collating apparatus can be realized by a one-dimensional integration calculation process. For this reason, a higher processing speed can be obtained.
[0027]
[Fine correction of mapped image (step A4)]
The image mapped by the above-described procedure has a slight shift in details, particularly when alignment is performed using ruled lines that are separated from each other. Therefore, in this image collating apparatus, the image after mapping is divided into small areas, and the small area is moved within a range of several pixels in the vertical and horizontal directions as shown in FIG.
[0028]
Conventionally, when this processing is performed, an image is divided into rectangular small areas. However, when such division is performed, a failure may occur in an image in which many straight lines appear.
[0029]
FIG. 6 shows an example of an image in which this failure occurs. FIG. 6A shows the shape of a conventional small area. In the image of FIG. 6B, a linear pattern (bold line in the center horizontal direction) exists at the boundary of the small area of the image after mapping. To do. Here, if this rectangular small area can be moved only by one pixel up, down, left, and right, by moving each small area so as to most closely match the reference image of FIG. 6D, the image of FIG. can get. Therefore, a pattern that frequently appears in the input image, such as a straight line, is erased by this processing. For this reason, a line which is a clear difference between the input image and the reference image is not detected.
[0030]
Therefore, in this image collation apparatus, in order to avoid this obstacle, a hexagonal shape as shown in FIG. 7 is adopted as the shape of the small region. The hexagon is the largest polygon that can be laid flat, and there is no continuous linear pattern at the boundary when laid. Therefore, the linear pattern is not completely erased.
[0031]
[Comparison with Reference Image (Step A5)]
Finally, the image collating apparatus compares the image obtained by the above-described procedure with the reference image. At this time, since there is a slight pixel blur in the image read from the scanner or the like, the difference pixels around the common pixel are not regarded as difference pixels as shown in FIG. By performing this process, a clear difference image can be finally obtained.
[0032]
(Second Embodiment)
Next, a second embodiment will be described. FIG. 9 shows the flow of processing of the image collating apparatus of the second embodiment. The image collating apparatus according to the second embodiment performs processing using a total of three images, ie, two input images and one reference image. More specifically, for example, a blank form image is prepared as a reference image, while a completed form image and a modified form image are prepared as input images, and processing for detecting only this corrected portion is performed. Do.
[0033]
In the second embodiment, the image collation apparatus performs processing up to enlargement, reduction, and rotation for each of the two input images (step B1 to step B6), and then finely corrects the two images so that they match best. (Step B7) Finally, the difference image is output (Step B8). For this reason, it is possible to perform comparison after aligning two images including skew so as to coincide with the reference image.
[0034]
【The invention's effect】
As described above in detail, according to the present invention, it is possible to significantly reduce the processing time of image matching by detecting ruled lines as preprocessing, checking the correspondence, and enlarging / reducing / rotating the examined image. Is possible.
[0035]
In addition, by arranging the divided small regions in a hexagonal crystal pattern, it is possible to prevent erroneous erasure of the linear pattern that frequently appears in the input image.
[Brief description of the drawings]
FIG. 1 is a diagram showing a device configuration of an image collating apparatus according to an embodiment of the present invention.
FIG. 2 is a view showing a processing flow of the image collating apparatus according to the first embodiment;
FIG. 3 is a view showing a correspondence relationship between four ruled lines on the input image and four ruled lines on the reference image processed by the image collating apparatus according to the first embodiment;
FIG. 4 is a view showing coordinates of intersections of four ruled lines on the input image and four ruled lines on the reference image processed by the image collating apparatus according to the first embodiment;
FIG. 5 is a view for explaining fine correction of a mapped image of the image collating apparatus according to the first embodiment;
FIG. 6 is a diagram for explaining a problem in fine correction of a mapped image of the image collating apparatus according to the first embodiment.
FIG. 7 is a diagram for explaining a problem in fine correction of a mapped image of the image collating apparatus according to the first embodiment.
FIG. 8 is a view for explaining comparison in which the difference pixels around the common pixel are ignored in the image collating apparatus according to the first embodiment;
FIG. 9 is a view showing a flow of processing of the image collating apparatus of the second embodiment.
[Explanation of symbols]
1 ... CPU
2 ... System memory 3 ... I / O controller 11 ... Image collation system

Claims

In an image collation apparatus that compares a reference image given in advance with a collated image that is input as needed to determine whether the reference image and the collated image match or not,
Ruled line detection means for detecting a ruled line from each of the reference image and the image to be verified;
Investigation means for investigating the correspondence between the reference image of the ruled line detected by the ruled line detection means and the collated image;
First correction means for performing deformation including any of enlargement, reduction, and rotation on the image to be matched so as to match the reference image from the correspondence relationship of the ruled lines obtained by the investigation means;
A second image in which the image to be verified corrected by the first correction means is divided into a plurality of hexagonal small areas and each of the small areas is moved within a predetermined range to match the reference image. Correction means,
Difference acquisition means for acquiring a difference between the image to be verified and the reference image after correction by the second correction means;
An image collating apparatus comprising:

In the image collation apparatus that compares the first and second images to be collated, which are input as needed, and determines whether or not they match,
A ruled line detecting means for detecting a ruled line from each of the first and second images to be collated and a predetermined reference image;
Investigation means for investigating a correspondence relationship between the reference image of the ruled line detected by the ruled line detection means and the first and second matching images;
A first correction that applies a deformation including any of enlargement, reduction, and rotation to each of the first and second images to be matched to match the reference image based on the correspondence relationship of the ruled lines obtained by the investigation means. Means,
After the correction by the first correcting means, the first collated image is divided into a plurality of small areas having a predetermined shape, and the predetermined collated image is matched with the second collated image within the predetermined range. A second correction means for moving each small area;
Difference obtaining means for obtaining a difference between the first and second collated images after correction by the second correction means;
An image collating apparatus comprising:

It said second correction means, the image matching system according to claim 2, characterized in that an array of hexagonal Akirajo the sequence of the small region.

In an image collation apparatus that compares a reference image given in advance with a collated image that is input as needed to determine whether the reference image and the collated image match or not,
A correction unit that divides the image to be verified into a plurality of hexagonal small areas and moves each of the small areas within a predetermined range to match the reference image;
Difference acquisition means for acquiring a difference between the image to be verified and the reference image after correction by the correction means;
An image collating apparatus comprising:

In the image collation apparatus that compares the first and second images to be collated, which are input as needed, and determines whether or not they match,
Correction means for dividing the first collated image into a plurality of hexagonal small regions and moving each of the small regions within a predetermined range to match the second collated image;
Difference obtaining means for obtaining a difference between the first and second images to be verified after correction by the correction means;
An image collating apparatus comprising:

In an image collation method for comparing a reference image given in advance with a collated image that is input as needed to determine a match or mismatch between the reference image and the collated image,
Detecting a ruled line from each of the reference image and the collated image;
Investigating the correspondence between the reference image of the ruled line detected and the image to be matched;
Performing a deformation including any of enlargement, reduction and rotation on the image to be matched to match the reference image from the correspondence relationship of the investigated ruled lines;
Dividing the corrected image to be collated into a plurality of hexagonal small areas and moving each of the small areas within a predetermined range to match the reference image;
Obtaining a difference between the image to be verified and the reference image after the movement;
An image collating method comprising:

In the image collating method for comparing the first and second images to be collated, which are input as needed, and determining whether or not they match,
Detecting a ruled line from each of the first and second images to be verified and a reference image given in advance;
Investigating the correspondence between the reference image of the detected ruled line and the first and second images to be verified;
Applying deformation including any of enlargement, reduction, and rotation to each of the first and second matching images to match the reference image from the correspondence relationship of the investigated ruled lines;
After the deformation, the first collated image is divided into a plurality of small regions having a predetermined shape, and each of the small regions is moved within a predetermined range to match the second collated image. Step to
Obtaining the difference between the first and second collated images after the movement;
An image collating method comprising:

8. The image matching method according to claim 7, wherein the arrangement of the small regions is a hexagonal arrangement.

In an image collation method for comparing a reference image given in advance with a collated image that is input as needed to determine a match or mismatch between the reference image and the collated image,
Dividing the image to be collated into a plurality of hexagonal small regions and moving each of the small regions within a predetermined range to match the reference image;
Obtaining a difference between the image to be verified and the reference image after the movement;
An image collating method comprising:

In the image collating method for comparing the first and second images to be collated, which are input as needed, and determining whether or not they match,
Dividing the first collated image into a plurality of hexagonal small regions and moving each of the small regions within a predetermined range to match the second collated image;
Obtaining the difference between the first and second collated images after the movement;
An image collating method comprising: