JPH0721382A - Picture processor - Google Patents
Picture processorInfo
- Publication number
- JPH0721382A JPH0721382A JP5165330A JP16533093A JPH0721382A JP H0721382 A JPH0721382 A JP H0721382A JP 5165330 A JP5165330 A JP 5165330A JP 16533093 A JP16533093 A JP 16533093A JP H0721382 A JPH0721382 A JP H0721382A
- Authority
- JP
- Japan
- Prior art keywords
- correlation value
- function
- calculation
- pixel
- correlation
- 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
Links
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Description
【０００１】[0001]
【産業上の利用分野】本発明は、画像処理装置に関し、
特に画像パターン認識分野において、或る画像パターン
を、検査対象画像データの中から探し出す場合に使用さ
れる。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus,
In particular, in the field of image pattern recognition, it is used when a certain image pattern is searched for in the inspection target image data.
【０００２】[0002]
【従来の技術】上記のような画像パターン認識分野にお
いて、或る画像パターン（リファレンス用）を、検査対
象画像データの中から探し出すとき、従来は、リファレ
ンス用のパターン（データ）と検査対象画像データすべ
ての位置（すべての画素部）で正規化相関計算を行い、
これにより相関値を求めて、これら相関値のうちの最大
相関値が得られた位置をリファレンス用パターンと一致
する位置としていた。従って、正規化相関計算をすべて
の位置で行う分、位置検出速度が遅くなるという欠点が
あった。2. Description of the Related Art In the field of image pattern recognition as described above, when a certain image pattern (for reference) is searched from the inspection target image data, conventionally, the reference pattern (data) and the inspection target image data are used. Normalized correlation calculation is performed at all positions (all pixel parts),
Thus, the correlation value is obtained, and the position where the maximum correlation value among these correlation values is obtained is set as the position that matches the reference pattern. Therefore, there is a drawback that the position detection speed becomes slower because the normalized correlation calculation is performed at all positions.
【０００３】図５は、この欠点を説明するための一画面
分の検査対象画像データを示す。ここで、１は画素部を
示し、２は正規化相関計算を行う位置、３は最大相関値
の求まった位置を示す。FIG. 5 shows inspection target image data for one screen for explaining this drawback. Here, 1 indicates a pixel portion, 2 indicates a position where a normalized correlation calculation is performed, and 3 indicates a position where the maximum correlation value is obtained.
【０００４】この図５を見て分かる通り、最大相関値３
を探し出すためには、正規化相関計算をすべての位置
２、３で行う必要があり、計算時間が著しく長くなって
しまうものであった。As can be seen from FIG. 5, the maximum correlation value 3
In order to find out, it is necessary to perform the normalized correlation calculation at all positions 2 and 3, and the calculation time becomes extremely long.
【０００５】[0005]
【発明が解決しようとする課題】本発明は上記実情に鑑
みてなされたもので、最大相関値がある画素部の位置
を、迅速に求めることができる画像処理装置を提供しよ
うとするものである。SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide an image processing apparatus capable of quickly obtaining the position of a pixel portion having the maximum correlation value. .
【０００６】[0006]
【課題を解決するための手段と作用】本発明は、画面を
構成するように並ぶ画素部のうち、市松模様の一方の色
配置の部分の画素部をアドレス指定する第１の機能、お
よび選択された画素部付近の一定範囲内にある全画素部
についての実質的なアドレス指定が行われるようにする
第２の機能を有するアドレス指定部と、このアドレス指
定部で指定された画素部の画像データとリファレンス用
パターンデータの相関値を正規化相関計算により求める
計算部とを具備し、前記計算部は、前記第１の機能を用
いて、前記市松模様の一方の色配置の部分の画素部につ
いての前記相関値を求め、これら相関値のうちの最大相
関値を得た画素部付近の一定範囲内の全画素部での相関
値を、前記第２の機能を用いて求め、この第２の機能で
求めた相関値のうちの最大相関値を得た画素部に対応す
る位置の検出を行うことを特徴とする。The present invention has a first function of addressing a pixel portion of one color arrangement of a checkered pattern among pixel portions arranged so as to form a screen, and selection. And an image of the pixel section designated by the addressing section, which has a second function for performing substantial addressing of all the pixel sections within a certain range near the specified pixel section. And a calculation unit that obtains a correlation value between the data and the reference pattern data by a normalized correlation calculation, wherein the calculation unit uses the first function, and the pixel unit of one color arrangement portion of the checkerboard pattern. Is obtained by using the second function, and the correlation values in all pixel portions within a certain range near the pixel portion that has obtained the maximum correlation value among these correlation values are obtained by using the second function. Correlation value calculated by the function And performing detection at a position corresponding to the pixel portion to obtain a maximum correlation value.
【０００７】すなわち本発明は、前記市松模様の一方の
色配置の部分の画素部についての相関値を求め、つまり
画面上の全ての画素部についての位置ではなく、とびと
びの位置についての相関値を求めて、それら相関値の中
から最大の相関値を求め、その最大相関値があった位置
の付近の全位置についてのみ集中的に相関値を求めて、
その中の最大の相関値があった位置を、求めたい位置と
する。即ち上記リファレンス用パターンデータに似たパ
ターンがある位置とする。従って、上記とびとびの相関
値を求めることによって、相関値計算を減らした分に近
い時間だけ、位置検出用の処理速度が速くできるもので
ある。That is, according to the present invention, the correlation value for the pixel portion of one color arrangement portion of the checkered pattern is obtained, that is, not the position for all the pixel portions on the screen, but the correlation value for the discrete position. Obtained, the maximum correlation value is obtained from those correlation values, and the correlation value is intensively obtained only for all positions near the position where the maximum correlation value was,
The position having the maximum correlation value among them is the position to be obtained. That is, it is assumed that there is a pattern similar to the reference pattern data. Therefore, by obtaining the discrete correlation values, the processing speed for position detection can be increased by a time close to the amount of reduced correlation value calculation.
【０００８】[0008]
【実施例】以下図面を参照して本発明の一実施例を説明
する。図１は、この実施例の構成説明図である。図中１
１は、画面を構成するように並ぶ画素部のうち、市松模
様の一方の色配置の部分の画素部をアドレス指定する第
１の機能、および選択された画素部付近の一定範囲内に
ある全画素部についての実質的なアドレス指定を行わせ
る第２の機能を有するアドレス指定部である。１２は、
このアドレス指定部１１で指定された画素部の画像デー
タａとリファレンス用パターンデータｂの相関値を正規
化相関計算により求める計算部である。この計算部１２
は、前記第１の機能を用いて、前記市松模様の一方の色
配置の部分の画素部についての前記相関値を求め、これ
ら相関値のうちの最大相関値を得た画素部付近の一定範
囲内の全画素部での相関値を、前記第２の機能を用いて
求め、この第２の機能で求めた相関値のうちの最大相関
値を得た画素部に対応する位置の検出を行う。１３は、
上記各画素部のデータａを蓄えたメモリである。このメ
モリ１３が、アドレス指定部１１によってアドレス指定
されるときは、そのアドレスを計算部１２にも知らせ、
計算部１２が、入力されたデータａのアドレスを判別可
能としている。また、計算部１２は、上記第１の機能を
用いて知った前者の最大相関値の位置情報を、アドレス
指定部１１に知らせるようにしている。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is an explanatory diagram of the configuration of this embodiment. 1 in the figure
1 is the first function of addressing the pixel portion of one color arrangement portion of the checkered pattern among the pixel portions arranged so as to form the screen, and all within a certain range near the selected pixel portion. It is an addressing unit having a second function for performing substantial addressing of the pixel section. 12 is
This is a calculation unit for obtaining the correlation value between the image data a of the pixel portion designated by the address designation unit 11 and the reference pattern data b by the normalized correlation calculation. This calculator 12
Using the first function, obtains the correlation value for the pixel portion of the one color arrangement portion of the checkerboard pattern, and obtains the maximum correlation value of these correlation values within a certain range near the pixel portion. The correlation values of all the pixel portions in the inside are obtained by using the second function, and the position corresponding to the pixel portion having the maximum correlation value among the correlation values obtained by the second function is detected. . 13 is
It is a memory that stores the data a of each pixel portion. When this memory 13 is addressed by the address designating section 11, it also informs the calculating section 12 of the address,
The calculator 12 can determine the address of the input data a. Further, the calculation unit 12 notifies the address designation unit 11 of the position information of the former maximum correlation value, which is obtained by using the first function.
【０００９】図２は、上記第１の機能を説明する場合の
１画面の検査対象画像データ配置を示す図、図３は、上
記第２の機能を説明する場合の１画面の検査対象画像デ
ータ配置を示す図である。図２において、２１は、画素
部を示し、２２は正規化相関計算を行う位置、２３は最
大相関値の求まった位置を示す。図３において、位置２
３は、図２により最大相関値の求まった位置で、この図
２のものと同様のものと考えてよい。２４は、上記第２
の機能により最大相関値の求めまった位置である。FIG. 2 is a diagram showing an arrangement of image data to be inspected on one screen when explaining the first function, and FIG. 3 is a diagram showing image data to be inspected on one screen when explaining the second function. It is a figure which shows arrangement. In FIG. 2, 21 indicates a pixel portion, 22 indicates a position where a normalized correlation calculation is performed, and 23 indicates a position where the maximum correlation value is obtained. In FIG. 3, position 2
Reference numeral 3 denotes a position where the maximum correlation value is obtained from FIG. 2, and may be considered to be the same as that in FIG. 24 is the second
This is the position where the maximum correlation value was obtained by the function of.
【００１０】次に、図１の構成を用いて、図２、図３に
ついて行う処理を、図４のフローチャートに沿って説明
する。まず、リファレンス用パターンと検査対象画像デ
ータとの間の輝度のずれをなくすため、これら両者間の
輝度の正規化を行う。（ステップＳ１） 次に、アドレス指定部１１による指示で取り込まれた図
２の検査対象画像データの左上端の画像データ（ａ）と
リファレンス用のパターンデータ（ｂ）の相関値を、計
算部１２での正規化相関計算により求める。（ステップ
Ｓ２） 次に、アドレス指定部１１の指示によって、図２の左上
端から右に２つずらした位置の画像データ（ａ）とパタ
ーンデータ（ｂ）の相関値を、計算部１２での正規化相
関計算により求める。（ステップＳ３） 次に、図２における検査対象画像データの右端になるま
で、ステップＳ３の処理を繰り返す。（ステップＳ４） 次に、ステップＳ４の処理で検査対象画像データの右端
まで来たならば、アドレス指定部１１の指示によって、
画面における計算対象の行（ライン）を１つ下げ、その
ライン＋１（１ライン上のスタート位置が、＋１右にず
れていた場合は左端）までずらし、その位置での画像デ
ータ（ａ）とパターンデータ（ｂ）の相関値を正規化相
関計算によって求める。（ステップＳ５） 次に、ステップＳ３〜Ｓ５の処理を繰り返しながら、検
査対象画像データの各位置２２での相関値を求める。
（ステップＳ６） 次に、ステップＳ２〜Ｓ６で求められた相関値より、最
大相関値を算出した位置２３を、計算部１２によって求
める。（ステップＳ７） 次に図３で示される如く、ステップＳ７で求められた位
置２３より上下左右にそれぞれ２つずらした範囲内での
すべての位置における相関値を求める。（ステップＳ
８） 次に、ステップＳ８で求めた範囲の相関値の中から、最
大の相関値を算出した位置２４を、リファレンス用パタ
ーンｂに最も似た画像データを持った位置であるとす
る。この時、位置２３と位置２４とが重なる場合がある
ことは勿論である。（ステップＳ９） 以上のように、図２の動作では、アドレス指定部１１に
よる指示で、検査対象画像データに対して１つ置きの位
置での相関値を、正規化計算により求め、その計算結果
から最大相関値を算出した位置を求める。次に図３の動
作により、図２の動作で求めた位置２３より上下左右に
例えば２×２の範囲だけ広げた位置の各位置での相関値
を、正規化相関計算によって求め、その計算結果の中で
最大の相関値を算出した位置２４を、検出位置とするも
のである。このような位置検出を行えば、その検出に要
する処理速度が、図５の従来例の場合よりも、３割〜４
割程度速くなった。Next, with reference to the flow chart of FIG. 4, the process of FIG. 2 and FIG. 3 will be described using the configuration of FIG. First, in order to eliminate the difference in brightness between the reference pattern and the image data to be inspected, the brightness between them is normalized. (Step S1) Next, the correlation value between the image data (a) at the upper left corner of the inspection target image data of FIG. It is obtained by the normalized correlation calculation in. (Step S2) Next, according to an instruction from the address designating section 11, the calculating section 12 obtains the correlation value between the image data (a) and the pattern data (b) at two positions shifted to the right from the upper left end in FIG. Calculated by normalized correlation calculation. (Step S3) Next, the process of step S3 is repeated until the right end of the inspection target image data in FIG. 2 is reached. (Step S4) Next, if the right end of the image data to be inspected is reached in the process of step S4,
The line (line) to be calculated on the screen is lowered by one, and the line is shifted to the line +1 (the left end if the start position on one line is shifted to the right by +1), and the image data (a) and the pattern at that position The correlation value of the data (b) is obtained by the normalized correlation calculation. (Step S5) Next, by repeating the processing of steps S3 to S5, the correlation value at each position 22 of the inspection target image data is obtained.
(Step S6) Next, the calculation unit 12 calculates the position 23 where the maximum correlation value is calculated from the correlation values calculated in steps S2 to S6. (Step S7) Next, as shown in FIG. 3, correlation values at all positions within the range shifted by two vertically and horizontally from the position 23 obtained at step S7 are obtained. (Step S
8) Next, from the correlation values in the range obtained in step S8, the position 24 where the maximum correlation value is calculated is assumed to be the position having the image data most similar to the reference pattern b. At this time, it goes without saying that the position 23 and the position 24 may overlap. (Step S9) As described above, in the operation of FIG. 2, the correlation value at every other position with respect to the image data to be inspected is obtained by the normalization calculation according to the instruction from the address designating unit 11, and the calculation result is obtained. Then, the position where the maximum correlation value is calculated is obtained. Next, by the operation of FIG. 3, the correlation value at each position of the position 23 obtained by the operation of FIG. The position 24 where the maximum correlation value is calculated is the detection position. If such position detection is performed, the processing speed required for the detection is 30% to 40% higher than that in the conventional example of FIG.
It's faster now.
【００１１】[0011]
【発明の効果】以上説明したごとく本発明によれば、正
規化相関計算の回数をほぼ半分に減少できることによ
り、画像パターン検出時の処理速度を、従来よりもかな
りの時間だけ速くすることができる。As described above, according to the present invention, the number of times of the normalized correlation calculation can be reduced to almost half, so that the processing speed at the time of detecting an image pattern can be increased by a considerable time as compared with the conventional case. .
【図１】本発明の一実施例の構成図。FIG. 1 is a configuration diagram of an embodiment of the present invention.
【図２】同実施例の動作説明に用いる１画面分の検査対
象画像データ配置を示す図。FIG. 2 is a diagram showing an arrangement of image data to be inspected for one screen, which is used for explaining the operation of the embodiment.
【図３】同実施例の動作説明に用いる１画面分の検査対
象画像データ配置を示す図。FIG. 3 is a diagram showing an arrangement of image data to be inspected for one screen, which is used for explaining the operation of the embodiment.
【図４】同実施例の動作説明に用いるフローチャート。FIG. 4 is a flowchart used to explain the operation of the embodiment.
【図５】従来例の動作説明に用いる１画面分の検査対象
画像データ配置を示す図。FIG. 5 is a diagram showing an inspection target image data arrangement for one screen used for explaining the operation of the conventional example.
１１…アドレス指定部、１２…計算部、１３…メモリ、
２１…画素部、２２…正規化相関計算を行う位置、２３
…第１の最大相関値の求まった位置、２４…第２の最大
相関値の求まった位置。11 ... Address designation unit, 12 ... Calculation unit, 13 ... Memory,
21 ... Pixel part, 22 ... Position for performing normalized correlation calculation, 23
... the position where the first maximum correlation value was found, 24 ... the position where the second maximum correlation value was found.
───────────────────────────────────────────────────── フロントページの続き (51)Int.Cl.6 識別記号 庁内整理番号 ＦＩ 技術表示箇所 Ｇ０６Ｔ 1/00 ─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. 6 Identification code Internal reference number FI technical display location G06T 1/00
Claims (2)
市松模様の一方の色配置の部分の画素部のデータが得ら
れるようなアドレス指示をする第１の機能、および選択
された画素部付近の一定範囲内にある全画素部のデータ
が得られるようなアドレス指示をする第２の機能を有す
るアドレス指定部と、このアドレス指定部で指示された
画素部の画像データとリファレンス用パターンデータの
相関値を正規化相関計算により求める計算部とを具備
し、前記計算部は、前記第１の機能を用いて、前記市松
模様の一方の色配置の部分の画素部についての前記相関
値を求め、これら相関値のうちの最大相関値を得た画素
部付近の一定範囲内の全画素部での相関値を、前記第２
の機能を用いて求め、この第２の機能で求めた相関値の
うちの最大相関値を得た画素部に対応する位置の検出を
行うことを特徴とする画像処理装置。1. A pixel unit arranged to form a screen,
The first function of addressing so that the data of the pixel portion of one color arrangement of the checkered pattern can be obtained, and the data of all pixel portions within a certain range near the selected pixel portion can be obtained. An address designating section having a second function for designating a different address, and a computing section for obtaining a correlation value between the image data of the pixel section designated by the address designating section and the reference pattern data by a normalized correlation calculation. The calculation unit uses the first function to obtain the correlation value for the pixel portion of one color arrangement portion of the checkered pattern, and obtains the maximum correlation value of these correlation values. The correlation value in all pixel parts within a certain range in the vicinity is calculated as the second value.
The image processing apparatus is characterized in that the position corresponding to the pixel portion that has obtained the maximum correlation value among the correlation values obtained by the second function is obtained using the above function.
数ビットである請求項１に記載の画像処理装置。2. The image processing apparatus according to claim 1, wherein the image data of each pixel portion has a plurality of bits.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP5165330A JPH0721382A (en) | 1993-07-05 | 1993-07-05 | Picture processor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP5165330A JPH0721382A (en) | 1993-07-05 | 1993-07-05 | Picture processor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH0721382A true JPH0721382A (en) | 1995-01-24 |
Family
ID=15810290
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP5165330A Pending JPH0721382A (en) | 1993-07-05 | 1993-07-05 | Picture processor |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH0721382A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008217206A (en) * | 2007-03-01 | 2008-09-18 | Fujifilm Corp | Image identification device, image identification method and imaging device |
| JP2012038276A (en) * | 2010-08-12 | 2012-02-23 | Casio Comput Co Ltd | Image identification device, image identification method and program |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS634379A (en) * | 1986-06-25 | 1988-01-09 | Toshiba Corp | Pattern matching device |
| JPH01159774A (en) * | 1987-12-17 | 1989-06-22 | Fujitsu Ltd | Image collating method and its circuit |
| JPH04293177A (en) * | 1991-03-22 | 1992-10-16 | Stanley Electric Co Ltd | Image processing method |
| JPH0540828A (en) * | 1991-08-05 | 1993-02-19 | Fujitsu Ltd | Block matching search system |
-
1993
- 1993-07-05 JP JP5165330A patent/JPH0721382A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS634379A (en) * | 1986-06-25 | 1988-01-09 | Toshiba Corp | Pattern matching device |
| JPH01159774A (en) * | 1987-12-17 | 1989-06-22 | Fujitsu Ltd | Image collating method and its circuit |
| JPH04293177A (en) * | 1991-03-22 | 1992-10-16 | Stanley Electric Co Ltd | Image processing method |
| JPH0540828A (en) * | 1991-08-05 | 1993-02-19 | Fujitsu Ltd | Block matching search system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008217206A (en) * | 2007-03-01 | 2008-09-18 | Fujifilm Corp | Image identification device, image identification method and imaging device |
| JP2012038276A (en) * | 2010-08-12 | 2012-02-23 | Casio Comput Co Ltd | Image identification device, image identification method and program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JPH0721382A (en) | Picture processor | |
| JP3796356B2 (en) | Camera control apparatus, method, and computer-readable storage medium | |
| US4656468A (en) | Pattern data processing apparatus | |
| JP3127598B2 (en) | Method for extracting density-varying constituent pixels in image and method for determining density-fluctuation block | |
| JPH0721383A (en) | Picture processor | |
| JP3483751B2 (en) | Motion vector detecting device and motion vector detecting method | |
| JPH0795469A (en) | Picture compensation device of camcorder | |
| JP3047952B2 (en) | Image processing device | |
| JP3717088B2 (en) | Appearance inspection method | |
| JP3437087B2 (en) | Feature point tracking method and apparatus, and recording medium storing feature point tracking program | |
| JP2833137B2 (en) | Pattern matching method | |
| JPH0214387A (en) | Plural peak points detecting system | |
| JPH04293177A (en) | Image processing method | |
| JPH05143733A (en) | Contour extracting device | |
| JPH07210689A (en) | Method for detecting picture position | |
| JPS6375989A (en) | Picture recognizing system | |
| JP2634905B2 (en) | How to paint shapes | |
| JPH10162138A (en) | Image processing method | |
| JP3096123B2 (en) | Document processing apparatus and method | |
| JP2900707B2 (en) | How to display design rule errors | |
| JPH05114028A (en) | Method and circuit for matching density pattern | |
| JP2000182056A (en) | Picture processor | |
| JPH10240869A (en) | Device for preparing character recognition dictionary and method therefor | |
| JPH08241419A (en) | Graphic plotting control method | |
| JPH0523464B2 (en) |