JP2004237532A - Detecting method and device of erroneous plate imposing - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To direct to a suitable technique for detecting the erroneous imposing of plates to be installed on a plate cylinder and realize the low cost and correct detection of the erroneous imposing of the plates. <P>SOLUTION: The positional deviation of the gradation level value f(x, y) of an image to be inspected is corrected based on the positional deviation amount (Δx, Δy) to the gradation level value f<SB>base</SB>(x, y) of a base image serving as an inspection standard. The number of pixels, the gradation level difference of the corrected image obtained by subtracting a secondarily differentiated value multiplied by coefficients corresponding to the positional deviation amount of the gradation level value f<SB>base</SB>(x, y) of the base image from a difference ¾f<SB>base</SB>(x, y)-F(x, y)¾ or the difference between the gradation level value F(x, y) of a corrected image, the positional deviation of which is corrected, and the gradation level value f<SB>base</SB>(x, y) of the base image of each of which is larger than the pre-set base level difference, is calculated. When the ratio of the calculated number of pixels to the total number of pixels is larger than the pre-set base ratio, the plates are detected to be erroneously imposed. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

【００１８】
このように、変換手段４１は、ＣＭＹＫの色情報として保存されている基準画像ｆ_ｂａｓｅ（ｘ，ｙ）を、ラインセンサカメラ３により取り込まれる被検査画像ｆ（ｘ，ｙ）と同じ形式、即ちＲＧＢＩｒ輝度値に数値化するようになっている。
【００１９】
位置ずれ算出手段４２は、変換手段４１により輝度値変換された基準画像ｆ_{ｂａ ｓｅ}（ｘ，ｙ）と被検査画像ｆ（ｘ，ｙ）との位置ずれ量（Δｘ，Δｙ）を、正規化相関法を用いて算出するようになっている。
【００２０】
正規化相関法では、図２に示すように、まず、テンプレートとして用意した２次元画像Ｔと、テンプレート画像Ｔと同じサイズの部分画像Ｉ′を対象画像Ｉから切り出し、これら２次元画像Ｔ，Ｉ′をそれぞれ１次元のベクトルと見なし、以下に示す式２により相関値Ｃを計算する。
【００２１】
【数２】

【００２２】
そして、この式２により、部分画像Ｉ′を１画素ずつずらしながら対象画像の全体について相関値Ｃの計算を行ない、相関値Ｃが最大となる点を、テンプレート画像Ｔが存在する点とする。このようにして、テンプレート画像Ｔと対象画像Ｉとの位置関係、即ち位置ずれを求めることができる。
【００２３】
本実施形態に係る位置ずれ算出手段４２は、基準画像（例えば画像全体：横１６０画素×縦２４０画素）の中央部分（例えば横１４０画素×縦２２０画素）をテンプレート画像Ｔとして抽出し、上述したような正規化相関法の計算を行なうことで、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）と被検査画像ｆ（ｘ，ｙ）との位置ずれ量（Δｘ，Δｙ）を算出するようになっている。
そして、最大相関値Ｃがあらかじめ設定された基準の相関値（基準相関値）よりも小さい場合、後述する版掛け間違い検出手段４４により、版胴６に装着された版５は、本来装着されるべき版５とは全く異なるものである（版の掛け間違いである）と検出されるようになっている。
【００２４】
位置ずれ補正手段４３は、位置ずれ算出手段４２により算出された位置ずれ量（Δｘ，Δｙ）に基づいて被検査画像ｆ（ｘ，ｙ）の位置ずれを補正するようになっている。
【００２５】
この位置ずれ補正について具体的に説明する。また、説明を段階的に行なうため、画像を１次元のものとする。つまり、基準画像をｆ_ｂａｓｅ（ｘ）、被検査画像をｆ（ｘ）として説明する。
図３に示すように、検査対象が、基準画像ｆ_ｂａｓｅ（ｘ）に対してｘ軸の正方向にΔｘだけずれている時、取り込まれる被検査画像ｆ（ｘ）は印刷欠陥（インキのぼた落ちや色抜け）がなければ式３で表わされる。
【００２６】
【数３】

【００２７】
また、被検査画像ｆ（ｘ）を位置ずれ補正した後の補正画像Ｆ（ｘ）は式４で表わされる。
【００２８】
【数４】

【００２９】
この式４により、被検査画像ｆ（ｘ）の位置ずれを補正することで補正画像の輝度値Ｆ（ｘ）を得ることができる。
また、ラインセンサカメラ３で取り込まれた被検査画像が基準画像よりも全体的に明るかったり暗かったりする場合があるので、このレベル差をなくすため、位置ずれ補正手段４３は、基準画像と補正画像との輝度値レベルをそろえるようになっている（スケーリングを行なう）。
【００３０】
このスケーリングの手順としては、まず、基準画像ｆ_ｂａｓｅ（ｘ）の輝度値の最大値Ｍａｘ_ｓｔｄ及び最小値Ｍｉｎ_ｓｔｄを求める（手順１）。次に、補正画像Ｆ（ｘ）の輝度値の最大値Ｍａｘ及び最小値Ｍｉｎを求める（手順２）。そして、以下の式５により、補正画像Ｆ（ｘ）の全画素について輝度値を変換する（手順３）。なお、式５において、νは補正画像の各画素の輝度値を示している。
【００３１】
【数５】

【００３２】
このような手順により、基準画像ｆ_ｂａｓｅ（ｘ）と補正画像Ｆ（ｘ）との輝度値レベルのスケーリングが行なわれるようになっている。
版掛け間違い検出手段４４は、まず、以下に示す式６により、補正画像Ｆ（ｘ）と基準画像ｆ_ｂａｓｅ（ｘ）との差から、基準画像ｆ_ｂａｓｅ（ｘ）の位置ずれ量に対応した係数を乗じた２次微分値を減算するようになっている。
【００３３】
【数６】

【００３４】
次に、版掛け間違い検出手段４４は、式６により得られた被検査画像ｆ（ｘ）の輝度値の差（濃淡レベル差）Ｓを予め設定された基準レベル差と比較する。
そして、Ｓが基準レベル差よりも大きくなる画素の数Ｎを計算し、全画素数Ｎ_０に対するＮの割合（即ち面積）Ｍを求め、このＭが予め設定された基準割合よりも小さい場合、版胴６に装着された版５は適正（正常）に装着されていると判定し、また、Ｍが予め設定された基準割合よりも大きい場合、版胴６に装着された版５は掛け間違いであると判定するようになっている。
【００３５】
ところで、上記の式６は、以下のような計算により求められたものである。まず、式４に式３を代入して以下の式７を得る。
【００３６】
【数７】

【００３７】
ここで、Δｘ（１−Δｘ）をｋとおいて整理すると、以下の式８が得られる。
【００３８】
【数８】

【００３９】
また、式８の右辺の中括弧の中を整理すると、以下の式９が得られる。
【００４０】
【数９】

【００４１】
この式９は、基準画像ｆ_ｂａｓｅ（ｘ）の差分の差分、すなわち基準画像ｆ_ｂａｓｅ（ｘ）の２次微分を示している。
さらに、画像を２次元で表示すると、即ち、基準画像をｆ_ｂａｓｅ（ｘ，ｙ）、被検査画像をｆ（ｘ，ｙ）とすると、上述と同様の考え方で、以下の式１０を得る。
【００４２】
【数１０】

【００４３】
なお、式１０の右辺第１項は基準画像ｆ_ｂａｓｅ（ｘ，ｙ）のｘ軸方向の２次微分値、右辺第２項は基準画像ｆ_ｂａｓｅ（ｘ，ｙ）のｙ軸方向の２次微分値、右辺第３項は基準画像ｆ_ｂａｓｅ（ｘ，ｙ）の斜め方向の２次微分値である。
上述したように、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）から補正画像Ｆ（ｘ，ｙ）を減算しても結果は０にはならず、ある値を持っている。この値は、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）の２次微分値と等しい値であることがわかる。
【００４４】
つまり、本検出装置では、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）から補正画像Ｆ（ｘ，ｙ）を減算した値から、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）の２次微分値をさらに減算することにより、より精度良く位置ずれを補正した上で、基準画像ｆ_ｂａｓｅ（ｘ，ｙ）と被検査画像ｆ（ｘ，ｙ）とのより詳細な違いを検出し、版５の掛け間違いが発生しているか否かをより正確に検出できるようになっている。
【００４５】
また、本検出装置では、版掛け間違い検出手段４４により版５の掛け間違いが検出された場合、表示手段としてのディスプレイ１０に、版５の掛け間違いが発生していることが表示されるようになっている。これにより、作業者は、ディスプレイ１０を見ることで、版５が掛け間違いであるか否かを容易に知ることができる。
【００４６】
本実施形態の版掛け間違い検出装置は、上述のように構成されているので、例えば図４に示すフローに沿って処理が行なわれる。
まず、製版システムに保存されている製版データ（基準画像）（ＣＭＹＫ，１ｂｉｔ ＴＩＦＦデータ）から、センサの各画素の画線率を計算する（ステップＳ１０。変換ステップ）。
【００４７】
次に、濃度変換テーブルＬＵＴを用いて各画素の画線率から濃度値を求め、さらに、濃度値から輝度値（ＲＧＢＩｒ）へ変換することにより、基準画像を、ラインセンサカメラ３により取り込まれた被検査画像のデータ形式（ＲＧＢＩｒ）と同様のデータ形式にする（ステップＳ２０。変換ステップ）。
その後、基準画像の中央部をテンプレートとして正規化相関法により基準画像に対する被検査画像の位置ずれ量を算出する（ステップＳ３０。位置ずれ算出ステップ）。なお、このとき、正規化相関法により算出された最大相関値が予め設定された基準相関値よりも小さいか否かを判定して（ステップＳ４０）、最大相関値が基準相関値よりも小さい場合、現在装着されている版は、本来装着されるべき版とは全く異なる版であると検出され（ステップＳ１１０）、ディスプレイ１０にその旨が表示される。
【００４８】
次に、被検査画像の位置ずれを補正して補正画像を求めた後（ステップＳ５０。位置ずれ補正ステップ）、この補正画像の輝度値レベルを基準画像の輝度値レベルに合わせてスケーリングを行なう（ステップＳ６０）。
その後、上記の式６の計算を行ない（ステップＳ７０）、式６により得られたＳが基準レベル差よりも大きくなる画素の数Ｎを求め、この画素数Ｎの全画素数Ｎ_０に対する割合Ｍを計算する（ステップＳ８０）。
【００４９】
そして、このＭが基準割合よりも大きい場合、現在装着されている版は、本来装着されるべき版とは部分的に異なる版或いは版全体が異なるものであると検出され（ステップＳ１２０。版掛け間違い検出ステップ）、ディスプレイ１０にその旨が表示される。
また、Ｍが基準割合以下である場合、現在装着されている版は本来装着されるべき版と同様であることがディスプレイ１０に表示される（ステップＳ１００）。
【００５０】
上述したように、本発明の一実施形態としての版掛け間違い検出装置及び検出方法によれば、正規化相関法により算出された基準画像と被検査画像との最大相関値に基づいて、版の掛け間違いが発生しているか否かをおおまかに判定することができる。
また、この後、基準画像と被検査画像との違いをさらに細かく調べることで、版の掛け間違いをより正確に検出することが可能である。
さらに、従来のようなバーコードリーダ等を版胴毎に設ける必要がないので、コストを低減できる。
【００５１】
以上、本発明の実施形態について説明したが、本発明は上記の実施形態に限定されるものではなく、本発明の趣旨を逸脱しない範囲で種々変形して実施することができる。
例えば、本実施形態では図１に示すように、本検出装置を製版システムとは別体に設ける構成としたが、本検出装置を製版システムと一体に構成してももちろん良い。
【００５２】
また、本実施形態では、図１に示すように片面印刷の場合について説明したが、両面印刷の場合には、紙面の両面側にラインセンサカメラ３を設けることで、紙面の両面側の版胴に装着される版についても、掛け間違いを検出することが可能であることは言うまでもない。
さらに、本発明は、輪転機だけでなく、刷版を用いるものに広く適用可能である。
【００５３】
【発明の効果】
以上詳述したように、本発明の版掛け間違い検出方法及び検出装置によれば、基準画像の濃淡レベル値ｆ_ｂａｓｅ（ｘ，ｙ）から補正画像の濃淡レベル値Ｆ（ｘ，ｙ）を減算した値から、基準画像の濃淡レベル値ｆ_ｂａｓｅ（ｘ，ｙ）の位置ずれ量に対応した係数を乗じた２次微分値をさらに減算することにより得られた補正画像の濃淡レベル差が、予め設定された基準レベル差よりも大きくなる画素の数を計算し、この画素数の全画素数に対する割合が基準割合よりも大きい場合に、版の掛け間違いとして検出するので、従来のようなバーコードリーダ等を版胴毎に設けて版の掛け間違いを判別する必要がなく、コストを低減できるとともに、より正確に版の掛け間違いを検出することが可能である。
【図面の簡単な説明】
【図１】本発明の一実施形態としての版掛け間違い検出装置を模式的に示す構成図（一部ブロック図）である。
【図２】本発明の一実施形態に係る正規化相関法を説明するための図である。
【図３】本発明の一実施形態に係る位置ずれ算出手段の処理を説明するための図である。
【図４】本発明の一実施形態としての版掛け間違い検出方法の処理を示すフローチャートである。
【図５】従来の版掛け間違い検出装置を模式的に示す構成図である。
【符号の説明】
１ 紙面
２ 絵柄
３ ラインセンサカメラ（画像取得手段）
３ａ 光源
４ 制御部（コンピュータ）
５ 版（刷版）
５ａ バーコード
６ 版胴
８ バーコードリーダ
１０ ディスプレイ（表示手段）
２０ 版掛け間違い判別装置
３０ 製版システム
４１ 変換手段
４２ 位置ずれ算出手段
４３ 位置ずれ補正手段
４４ 版掛け間違い検出手段[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a technique suitable for detecting an error in mounting a printing plate mounted on a plate cylinder, and particularly relates to a shading level between an image to be inspected printed on a printing surface and a reference image serving as an inspection reference. The present invention relates to a printing error detection method and a printing error detection device that perform detection based on a difference.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, in a printing machine (for example, a rotary press) which performs printing by mounting a printing plate on a plate cylinder, an operator manually mounts the printing plate on the plate cylinder before starting printing.
For this reason, when installing a printing plate on the plate cylinder, an operator may mistakenly mount the printing plate upside down or upside down, or install a printing plate that is different from the plate to be mounted. (This may be referred to as a plate misplacement or plate misplacement).
Therefore, the operator has visually inspected the printed matter actually printed, and has determined that the printing plate has been erroneously hooked.
However, in a rotary press or the like that prints a multi-page printed material such as newspaper, for example, the number of printing plates is large.
[0003]
Therefore, in order to save such labor of a worker, a printing press equipped with a device for automatically detecting a mistake in mounting a plate has been developed.
For example, in Patent Literature 1, as shown in FIG. 5, a bar code 5a for distinguishing the plate 5 is previously written in a left or right margin portion of the plate 5, and the A technique is disclosed in which the barcode 5a is read by a barcode reader (or camera) 8 to determine whether the plate 5 is correctly mounted.
Further, Patent Document 2 stores image data before printing plate production, and sets the image data at a position facing the peripheral surface of the plate cylinder at a predetermined time before printing is started after the printing plate is mounted on the plate cylinder. The image of the plate is read by the read sensor, and the image data is compared with the image data before the plate production. If the two data do not match, the rotation of the plate cylinder is stopped, and the plate is displayed on the display. A technique for displaying erroneous mounting is disclosed.
[0004]
[Patent Document 1]
Japanese Patent No. 3315061 [Patent Document 2]
Japanese Patent Publication No. 2971602
[Problems to be solved by the invention]
However, in the technique disclosed in Patent Document 1, it is necessary to write the barcode 5a on the plate 5, and a space (barcode space) for this needs to be secured.
Further, in the technique of Patent Document 1, it is necessary to equip all of the plate cylinders on which the printing plates are mounted with the barcode readers 8, resulting in high costs.
[0006]
Further, in the technique of Patent Literature 2, it is necessary to equip all of the plate cylinders on which the printing plates are mounted with the reading sensors, which leads to an increase in cost similarly to the technique of Patent Literature 1.
SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and has as its object to provide a printing error detection method and a printing error detection device which are capable of accurately detecting a printing error at low cost and accurately. I do.
[0007]
[Means for Solving the Problems]
For this reason, the plate imprinting error detecting method (Claim 1) of the present invention converts the inspection image printed on the printing surface into a gray level value f (x, y) using the position (x, y) of the printing surface as a parameter. y), and the displacement of the gray level f (x, y) of the image to be inspected with respect to the gray level f _base (x, y) of the reference image stored in advance as the inspection standard A displacement calculating step for calculating the amount (Δx, Δy), and a displacement correcting step for correcting the density level value f (x, y) of the inspection image based on the displacement (Δx, Δy). And the difference | f _base (x, y) -F (x, y) between the gray level value F (x, y) of the corrected image subjected to the positional deviation correction and the gray level value f _base (x, y) of the reference image. y) |, the gray level value f _base (x, y ) Is calculated by calculating the number of pixels in which the gray level difference of the corrected image obtained by subtracting the secondary differential value multiplied by the coefficient corresponding to the displacement amount is larger than a preset reference level difference, The present invention is characterized in that a printing error detection step is provided which detects a printing error when the ratio of the number of pixels to the total number of pixels is larger than a preset reference ratio.
[0008]
In the position shift calculating step, the position shift amount (Δx, y) of the gray level value f (x, y) of the inspection image with respect to the gray level value f _base (x, y) of the reference image is calculated by a normalized correlation method. Δy) is preferably calculated (claim 2).
If the maximum correlation value between the gray level value f _base (x, y) of the reference image and the gray level value f (x, y) of the image to be inspected is smaller than a predetermined reference correlation value, It is preferable to detect it as a wrong fit (claim 3).
[0009]
An imprinting error detecting device (claim 4) according to the present invention includes: an image acquiring unit that acquires an image to be inspected printed from a printing surface; and an image acquiring unit that acquires the inspected image acquired by the image acquiring unit. In addition to converting the position (x, y) into a gray level value f (x, y) using a parameter as a parameter, a reference image serving as an inspection reference is converted into a gray level value f using the position (x, y) on the printing surface as a parameter. _base (x, y) conversion means for converting the position deviation amount of gray level values _f base (x, y) of the reference image gray level value obtaining step images for f (x, y) (Δx , Δy ), And a position for correcting the gray level f (x, y) of the image to be inspected based on the amount of positional deviation (Δx, Δy) calculated by the positional deviation calculating unit. A shift correcting unit, and the position shift correcting unit. Positional deviation corrected gray level value F (x, y) of the corrected image and the difference between the gray level value _f base of the reference image _{(x, y) | f base} (x, y) -F (x, y) | Is subtracted from the secondary differential value obtained by multiplying the gray level value f _base (x, y) of the reference image by a coefficient corresponding to the amount of displacement, the gray level difference of the image to be inspected is A printing error detecting means for calculating the number of pixels larger than a predetermined reference level difference and detecting as a printing error when the ratio of the number of pixels to the total number of pixels is larger than a predetermined reference ratio; It is characterized by having.
[0010]
The displacement calculator calculates the displacement (Δx, Δy) of the gray level f (x, y) of the inspection image with respect to the gray level f _base (x, y) of the reference image by a normalized correlation method. The printing error detection means calculates the maximum correlation value between the gray level value f _base (x, y) of the reference image and the gray level value f (x, y) of the image to be inspected in advance. If the reference correlation value is smaller than the reference correlation value, it is preferable to detect that the plate is incorrectly attached (claim 5).
[0011]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic configuration diagram schematically showing a plate improper detection device as one embodiment of the present invention.
As shown in FIG. 1, a printing error detecting apparatus according to the present embodiment is provided, for example, in a rotary press for printing on continuous paper (paper surface) 1, and includes a line sensor camera 3 as an image acquisition unit, It comprises a control unit 4 such as a computer connected to the sensor camera 3 and a display 10 as display means.
[0012]
The line sensor camera 3 includes a CCD image sensor element, a lens, and the like (not shown). The same pattern 2 is detected by detecting the amount of reflected light from the paper surface 1 illuminated by a light source (for example, a fluorescent lamp) 3a. The image of the paper 1 printed continuously is taken in. Further, here, the line sensor camera 3 converts an image to be inspected (hereinafter, referred to as an image to be inspected) for detecting a misplacement of a plate with a gray level value of RGBIr (Red, Green, Blue, Infrared). (Light amount value or luminance value).
[0013]
The control unit 4 is provided with functions corresponding to the conversion unit 41, the position shift calculating unit 42, the position shift correcting unit 43, and the imprinting error detecting unit 44.
The conversion unit 41 converts the inspection image captured by the line sensor camera 3 into an x-axis in the width direction of the paper surface 1 and a y-axis in the traveling direction of the paper surface 1, that is, the position (x, y) of the printing surface as a parameter. Is converted into image data as a gray level value (specifically, a luminance value) f (x, y).
[0014]
The rotary press according to the present embodiment is connected to a plate making system 30 for managing plate making information by computer, and the plate making system 30 stores plate making data (reference image) of the picture 2 to be provided for printing. It is input and stored in advance. Here, the reference image is stored in CMYK (Cyan, Magenta, Yellow, blackK) 200 dpi, 1 bit / dot data [1 bit TIFF (Tagged Image File Format)].
[0015]
The conversion means 41 fetches a reference image of 1 dpi / dot data of 200 dpi from the plate making system 30 described above, and from this reference image, the image ratio of each pixel of the sensor (for example, the dot with ink in one pixel is represented by " When "1" is set and the other dots are set to "0", the ratio of "1" in one pixel) is calculated.
[0016]
The conversion unit 41 converts the CMYK image ratio of the reference image into a density value using a density conversion table (LUT: Look Up Table, also referred to as CMYK image ratio-density value table) prepared in advance. The density value is converted into an RGBIr luminance value by using Expression 1 shown in FIG. In Equation 1, G is the reference brightness (the brightness value of a blank sheet), and g is the brightness value.
[0017]
(Equation 1)

[0018]
As described above, the conversion unit 41 converts the reference image f _base (x, y) stored as the CMYK color information into the same format as the image to be inspected f (x, y) captured by the line sensor camera 3, that is, The RGBIr luminance value is digitized.
[0019]
Positional deviation calculating means 42, the reference image _{f ba} se (x, y) which has been converted luminance value by the conversion means 41 and the inspection image f (x, y) position shift amount between the ([Delta] x, [Delta] y), normalized The calculation is performed using a correlation method.
[0020]
In the normalized correlation method, as shown in FIG. 2, first, a two-dimensional image T prepared as a template and a partial image I ′ having the same size as the template image T are cut out from the target image I, and these two-dimensional images T, I 'Are each regarded as a one-dimensional vector, and a correlation value C is calculated by the following equation 2.
[0021]
(Equation 2)

[0022]
Then, the correlation value C is calculated for the entire target image while shifting the partial image I ′ one pixel at a time by Expression 2, and the point where the correlation value C becomes the maximum is defined as the point where the template image T exists. In this manner, the positional relationship between the template image T and the target image I, that is, the positional deviation can be obtained.
[0023]
The displacement calculating unit 42 according to the present embodiment extracts a central portion (for example, 140 pixels wide × 220 pixels vertically) of the reference image (for example, the entire image: 160 pixels wide × 240 pixels vertically) as the template image T, and as described above. By performing the calculation using the normalized correlation method, the amount of displacement (Δx, Δy) between the reference image f _base (x, y) and the image f (x, y) to be inspected is calculated. .
When the maximum correlation value C is smaller than a preset reference correlation value (reference correlation value), the printing plate 5 mounted on the printing drum 6 is originally mounted by the printing error detection unit 44 described later. It is to be detected that it is completely different from the power plate 5 (it is a mistake in mounting the plate).
[0024]
The position shift correcting unit 43 corrects the position shift of the inspection image f (x, y) based on the position shift amount (Δx, Δy) calculated by the position shift calculating unit 42.
[0025]
This misalignment correction will be specifically described. In addition, in order to provide a step-by-step description, the image is one-dimensional. That is, a description will be given assuming that the reference image is f _base (x) and the image to be inspected is f (x).
As shown in FIG. 3, when the inspection target is shifted by Δx in the positive x-axis direction with respect to the reference image f _base (x), the captured inspection image f (x) has a printing defect (ink blur). If there is no dropout or color loss, it is expressed by Equation 3.
[0026]
[Equation 3]

[0027]
Further, the corrected image F (x) after the positional deviation of the image to be inspected f (x) is corrected is expressed by Expression 4.
[0028]
(Equation 4)

[0029]
By using Equation 4, it is possible to obtain the luminance value F (x) of the corrected image by correcting the displacement of the image f (x) to be inspected.
In addition, since the image to be inspected captured by the line sensor camera 3 may be generally brighter or darker than the reference image, in order to eliminate this level difference, the displacement correction unit 43 uses the reference image and the corrected image. (The scaling is performed).
[0030]
As a procedure of this scaling, first, the maximum value Max _std and the minimum value Min _std of the luminance value of the reference image f _base (x) are obtained (procedure 1). Next, the maximum value Max and the minimum value Min of the luminance value of the corrected image F (x) are obtained (procedure 2). Then, the luminance value is converted for all the pixels of the corrected image F (x) according to the following Expression 5 (procedure 3). In Expression 5, ν indicates the luminance value of each pixel of the corrected image.
[0031]
(Equation 5)

[0032]
According to such a procedure, scaling of the luminance value level between the reference image f _base (x) and the corrected image F (x) is performed.
Edition hook inaccurate detection means 44, first, the equation 6 shown below, from the difference between the corrected image F (x) the reference image _{f base} (x), corresponding to the displacement amount of the reference image _{f base} (x) A second derivative value multiplied by a coefficient is subtracted.
[0033]
(Equation 6)

[0034]
Next, the imprinting error detecting means 44 compares the difference (shade level difference) S of the brightness value of the image to be inspected f (x) obtained by Equation 6 with a preset reference level difference.
Then, the number N of pixels in which S is greater than the reference level difference is calculated, and the ratio (ie, area) M of N to the total number of pixels N ₀ is obtained. If this M is smaller than a preset reference ratio, It is determined that the plate 5 mounted on the plate cylinder 6 is properly (normally) mounted. If M is larger than a preset reference ratio, the plate 5 mounted on the plate cylinder 6 is incorrectly mounted. Is determined.
[0035]
Incidentally, the above equation 6 is obtained by the following calculation. First, the following Expression 7 is obtained by substituting Expression 3 into Expression 4.
[0036]
(Equation 7)

[0037]
Here, by rearranging Δx (1−Δx) as k, the following Expression 8 is obtained.
[0038]
(Equation 8)

[0039]
Also, by rearranging the inside of the curly braces on the right side of Expression 8, the following Expression 9 is obtained.
[0040]
(Equation 9)

[0041]
This equation 9, the difference of the difference between the reference image _{f base} (x), that is, the second derivative of the reference image _{f base} (x).
Further, when the image is displayed in two dimensions, that is, when the reference image is f _base (x, y) and the image to be inspected is f (x, y), the following Expression 10 is obtained based on the same concept as described above.
[0042]
(Equation 10)

[0043]
Note that the first term on the right side of Expression 10 is the second derivative of the reference image f _base (x, y) in the x-axis direction, and the second term on the right side is the second order derivative of the reference image f _base (x, y) in the y-axis direction. The differential value and the third term on the right side are the secondary differential values in the oblique direction of the reference image f _base (x, y).
As described above, even if the corrected image F (x, y) is subtracted from the reference image f _base (x, y), the result does not become 0 but has a certain value. It can be seen that this value is equal to the second derivative of the reference image f _base (x, y).
[0044]
That is, in this detection device, the reference image _f base (x, y) correction image F (x, y) from the value obtained by subtracting the reference image _f base (x, y) to further subtract the second derivative of Then, after correcting the misalignment with higher accuracy, a more detailed difference between the reference image f _base (x, y) and the image f (x, y) to be inspected is detected, and a mistake in mounting the plate 5 occurs. It can be more accurately detected whether or not there is.
[0045]
Further, in the present detection apparatus, when the misappropriation of the plate 5 is detected by the misappropriation of the plate 5, the display 10 as the display means is displayed so as to indicate that the misappropriation of the plate 5 has occurred. Has become. Thus, the worker can easily know whether or not the plate 5 has been hung by looking at the display 10.
[0046]
Since the printing error detecting device of the present embodiment is configured as described above, the processing is performed, for example, according to the flow shown in FIG.
First, the image ratio of each pixel of the sensor is calculated from the plate making data (reference image) (CMYK, 1-bit TIFF data) stored in the plate making system (step S10: conversion step).
[0047]
Next, by using the density conversion table LUT, a density value is obtained from the image ratio of each pixel, and further converted from the density value to a luminance value (RGBIr), so that the reference image is captured by the line sensor camera 3. A data format similar to the data format (RGBIr) of the image to be inspected is used (step S20, conversion step).
Thereafter, the amount of displacement of the image to be inspected with respect to the reference image is calculated by the normalized correlation method using the central portion of the reference image as a template (step S30: displacement calculation step). At this time, it is determined whether or not the maximum correlation value calculated by the normalized correlation method is smaller than a preset reference correlation value (step S40), and if the maximum correlation value is smaller than the reference correlation value The currently mounted plate is detected as a plate completely different from the plate to be mounted (step S110), and the fact is displayed on the display 10.
[0048]
Next, after correcting the positional deviation of the image to be inspected to obtain a corrected image (Step S50: positional deviation correcting step), scaling is performed so that the luminance value level of the corrected image matches the luminance value level of the reference image (Step S50). Step S60).
Then, perform the calculation of the above equation 6 (step S70), obtains the number N of larger pixels than S is the reference level difference obtained by Equation 6, the ratio M relative to the total number of pixels N ₀ of the number of pixels N Is calculated (step S80).
[0049]
If this M is larger than the reference ratio, it is detected that the currently mounted plate is a plate partially different from the plate to be originally mounted or the whole plate is different (step S120: plate hanging). (Error detection step), and the fact is displayed on the display 10.
If M is equal to or less than the reference ratio, the display 10 indicates that the currently mounted plate is the same as the plate to be mounted (step S100).
[0050]
As described above, according to the printing error detection apparatus and the detection method according to an embodiment of the present invention, based on the maximum correlation value between the reference image and the inspection image calculated by the normalized correlation method, It can be roughly determined whether or not an error has occurred.
Further, after that, the difference between the reference image and the image to be inspected is examined more minutely, so that it is possible to more accurately detect an error in mounting the plate.
Further, since it is not necessary to provide a bar code reader or the like for each plate cylinder as in the related art, the cost can be reduced.
[0051]
Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and can be variously modified and implemented without departing from the gist of the present invention.
For example, in the present embodiment, as shown in FIG. 1, the present detection device is provided separately from the plate making system. However, the present detection device may be formed integrally with the plate making system.
[0052]
Also, in the present embodiment, the case of single-sided printing has been described as shown in FIG. 1, but in the case of double-sided printing, the line sensor cameras 3 are provided on both sides of the sheet to allow the plate cylinders on both sides of the sheet. Needless to say, it is possible to detect a mistake in mounting the plate mounted on the device.
Further, the present invention is widely applicable not only to rotary presses but also to those using printing plates.
[0053]
【The invention's effect】
As described above in detail, according to the printing error detecting method and the detecting device of the present invention, the gray level value F (x, y) of the corrected image is subtracted from the gray level value f _base (x, y) of the reference image. The gray level difference of the corrected image obtained by further subtracting a second derivative value multiplied by a coefficient corresponding to the positional deviation amount of the gray level value f _base (x, y) of the reference image from the obtained value is calculated in advance. Calculate the number of pixels that are larger than the set reference level difference, and if the ratio of this number of pixels to the total number of pixels is larger than the reference ratio, it will be detected as an incorrectly attached plate, There is no need to provide a reader or the like for each plate cylinder to determine whether a plate has been misplaced. This makes it possible to reduce costs and more accurately detect misplaced plates.
[Brief description of the drawings]
FIG. 1 is a configuration diagram (partially a block diagram) schematically showing an imprinting error detection device as one embodiment of the present invention.
FIG. 2 is a diagram for explaining a normalized correlation method according to an embodiment of the present invention.
FIG. 3 is a diagram for explaining processing of a displacement calculating unit according to an embodiment of the present invention.
FIG. 4 is a flowchart showing a process of a printing error detection method according to an embodiment of the present invention.
FIG. 5 is a configuration diagram schematically showing a conventional plate misplacement detection device.
[Explanation of symbols]
1 space 2 picture 3 line sensor camera (image acquisition means)
3a light source 4 control unit (computer)
5 plate (printing plate)
5a bar code 6 plate cylinder 8 bar code reader 10 display (display means)
Reference Signs List 20 plate misprinting discrimination device 30 plate making system 41 conversion means 42 misregistration calculation means 43 misregistration correction means 44 misplacing error detection means

Claims (5)

A conversion step of converting the inspection image printed on the printing surface into a gray level value f (x, y) using the position (x, y) of the printing surface as a parameter;
Calculate the positional deviation amount (Δx, Δy) of the gray level value f (x, y) of the image to be inspected with respect to the gray level value f _base (x, y) of the reference image which is stored in advance as the inspection standard. A displacement calculating step;
A displacement correction step of correcting a density level value f (x, y) of the inspection image based on the displacement amount (Δx, Δy);
Difference | f _base (x, y) -F (x, y) between the gray level value F (x, y) of the corrected image subjected to the positional deviation correction and the gray level value f _base (x, y) of the reference image | Is subtracted from the secondary differential value obtained by multiplying the gray level value f _base (x, y) of the reference image by a coefficient corresponding to the positional deviation amount, and the gray level difference of the corrected image is obtained in advance. Calculating a number of pixels larger than the set reference level difference, and detecting a plate mounting error when the ratio of the number of pixels to the total number of pixels is larger than a predetermined reference ratio; A printing error detection method, characterized in that: In the displacement calculation step,
Calculating the displacement (Δx, Δy) of the gray level f (x, y) of the inspection image with respect to the gray level f _base (x, y) of the reference image by the normalized correlation method. 2. The method according to claim 1, wherein the printing error is detected. If the maximum correlation value between the gray level value f _base (x, y) of the reference image and the gray level value f (x, y) of the image to be inspected is smaller than a predetermined reference correlation value, 3. The method according to claim 2, wherein the error is detected as an error. Image acquisition means for acquiring an image to be inspected printed from a printing surface,
The inspection image acquired by the image acquisition means is converted into a gray level value f (x, y) using the position (x, y) of the printing surface as a parameter, and a reference image serving as an inspection reference is converted to a reference image. Converting means for converting the print surface position (x, y) into a gray level value f _base (x, y) using the parameter as a parameter;
Position shift calculating means for calculating a position shift amount (Δx, Δy) of the gray level value f (x, y) of the inspection image with respect to the gray level value f _base (x, y) of the reference image; Misregistration correcting means for compensating for the density level value f (x, y) of the inspection image based on the misregistration amount (Δx, Δy) calculated by the calculating means;
The difference | f _base (x, y) − between the gray level value F (x, y) of the corrected image subjected to the positional deviation correction by the positional deviation correcting means and the gray level value f _base (x, y) of the reference image. The image to be inspected obtained by subtracting a secondary differential value obtained by multiplying a density level value f _base (x, y) of the reference image by a coefficient corresponding to the displacement amount from F (x, y) | Calculates the number of pixels whose gray level difference is larger than a preset reference level difference. If the ratio of the number of pixels to the total number of pixels is larger than a preset reference ratio, it is detected as an error in plate application. A printing error detecting device, comprising: a printing error detecting means. The position shift calculating means includes:
Calculating a positional shift amount (Δx, Δy) of the gray level value f (x, y) of the inspection image with respect to the gray level value f _base (x, y) of the reference image by a normalized correlation method;
The printing error detecting means includes:
If the maximum correlation value between the gray level value f _base (x, y) of the reference image and the gray level value f (x, y) of the image to be inspected is smaller than a predetermined reference correlation value, The printing error detecting device according to claim 4, wherein the error is detected as a mounting error.

Images