JP4028654B2 - Defect inspection equipment - Google Patents

Description

の場合が理想的な黒であることからして、適正な黒を取り込んでいることを表し
【外２】

に相当するレベルとなっており、出力信号Ｒ´としては何らホワイトバランス補正を必要としていないことを示している。なお、黒信号、白信号の平均値は、基準板の黒領域１１ｂ並びに白領域１１ｗにおいて一定の矩形領域が設定されており、この矩形領域内の平均輝度をいうものである。
【００１９】
【外３】

からずれていることになる。このため、黒信号は３０ほど下げるように補正し、白信号は５５ほど上げるように補正することにより、適正な色信号として取り込むことができる。ここでは、赤（Ｒ）のみを例に説明したが、同様な補正処理を他の緑（Ｇ）、青（Ｂ）についても行うことにより、各色信号が適正な信号として取り込むことができる。
【００２０】
以上の関係を数式で表現すると、次のようになる。ここで、Ｒは入力信号、
【外４】

平均レベル、Ｌｔは目標レベル（基準板の白領域がこのラベルに変換される）である。なお、上記は赤（Ｒ）の例であるが、式中、緑はＧ、青はＢであって、それぞれ上記の意味と同一である。
【００２１】
【数１】

【数２】

【数３】

上記各数式に各色毎の数値を代入してその結果を該当する濃度変換部に設定することによって、適切なホワイトバランス調整が行われる。
【００２２】
【発明の効果】
本発明によれば、検査対象が搭載される位置の近傍に白領域並びに黒領域からなる基準板を設け、検査対象を撮像すると同時にこの基準板の白領域並びに黒領域を撮像し、これら基準板からの信号に基づいてホワイトバランス調整を実施することができるようになり、このためリアルタイムでホワイトバランスの調整を行うことができるようになった。
【図面の簡単な説明】
【図１】本発明の一実施例を示す概略構成図である。
【図２】本発明の動作を説明するためのグラフである。
【図３】従来技術を示す概略構成図である。
【符号の説明】
１１ 基準板
１１ｗ 白領域
１１ｂ 黒領域
１２ 基準板画像入力部
１３ ホワイトバランスパラメータ算出処理部
３２ シート
３３ ３板型ＣＣＤラインセンサカメラ
３８Ｒ Ｒ−ＣＣＤ
３８Ｇ Ｇ−ＣＣＤ
３８Ｂ Ｂ−ＣＣＤ
３９Ｒ Ｒ画像入力部
３９Ｇ Ｇ画像入力部
３９Ｂ Ｂ画像入力部
４１Ｒ Ｒ濃度変換部
４１Ｇ Ｇ濃度変換部
４１Ｂ Ｂ濃度変換部
４３ 色変換部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a defect inspection apparatus using a color line sensor camera.
[0002]
[Prior art]
In recent years, inspecting the surface quality of various parts, the printing status of patterns, characters, etc., the quality of baking, painting, etc. on the surface of the material, the surface to be inspected is imaged with a camera, and the resulting image is processed. By doing so, apparatuses for optically inspecting defects have become widespread. Whether the inspection target is good or bad is determined with respect to dirt, scratches, colors, and the like on the surface, and the inspection is normally performed with the article to be inspected being moved. In particular, black-and-white images have been the mainstream of conventional captured images, but color inspection using a color camera has also become an inspection item, and in addition to the complexity of surface patterns and the demand for high-precision inspection specifications. There is a need for an inspection device. FIG. 3 shows an example of an inspection apparatus for inspecting a sheet surface defect using a three-plate CCD line sensor camera as an apparatus that meets such requirements.
[0003]
The inspection apparatus shown in FIG. 3 first picks up an image of the surface of the sheet 32 being conveyed in the direction of the arrow 31 as an inspection object by the three-plate CCD line sensor camera 33. Characters 34 and timing marks 35 printed in color are formed on the surface of the sheet 32. The three-plate CCD line sensor camera 33 includes a light receiving lens 36, a spectroscopic prism 37 arranged at a position where the optical axis is adjusted, and red (R), green (G), and blue separated by the spectroscopic prism 37. It has an R-CCD38R, G-CCD38G, and B-CCD38B that receive light components of the three colors (B). Each reflecting surface of the spectroscopic prism 37 is a band-pass filter, and reflects only the light components of RGB colors.
[0004]
Output signals from the CCDs 38R, 38G, and 38B are input to the R image input unit 39R, the G image input unit 39G, and the B image input unit 39B. A timing mark detection signal from a mark sensor 40 that detects a timing mark 35 formed on the sheet 32 is input to each of the R, G, and B image input units 39R, 39G, and 39B. As a trigger, it has a function of capturing image signals from the CCDs 38R, 38G, and 38B.
[0005]
In the case of precise color inspection, it is important to accurately capture the actual color to be inspected, but there may be variations in the output color signal depending on the characteristics of each CCD. As a result, a color different from the color to be inspected is output, and a correct inspection result cannot be obtained. For this reason, so-called white balance adjustment is performed. That is, the color reproducibility is improved by correcting the deviation from the RGB components constituting the standard color with respect to the output signal of each color. The output signals from the RGB image input units 39R, 39G, and 39B in FIG. 3 are an R density conversion unit 41R for red, a G density conversion unit 41G for green, and a B density conversion unit for blue. White balance adjustment is performed at 41B. This white balance adjustment is performed using a fixed white balance parameter set for each color component, and a value stored in the white balance parameter setting unit 42 is used as the fixed white balance parameter for each color.
[0006]
The signals of each color whose white balance has been adjusted are input to the color conversion unit 43. The color conversion unit 43 converts the RGB signal into a so-called HSI signal. The HSI signal represents hue (H), saturation (S), and lightness (I), and image processing for subsequent inspection is performed using this HSI signal. Each HSI signal is stored in the H frame memory 44H. , S frame memory 44S and I frame 44I, respectively. In addition, the captured image can be observed by the display device 45 using each color signal from the RGB density conversion units 41R, 41G, and 41B.
[0007]
[Problems to be solved by the invention]
However, in the conventional inspection apparatus, white balance adjustment is performed depending on a fixed white balance parameter. This is a method in which a white balance parameter specific to the camera is obtained in advance and this parameter is set before the camera is used for the first time. If various parts and circuits such as a CCD in the camera do not affect the white balance fluctuation over time, this fixing method is not a problem, but actually, various fluctuations occur over time. That is, the white balance situation changes due to fluctuations due to changes in the usage environment such as temperature, changes over time in various characteristics of camera components, and the like. For this reason, if the white balance of the image from the RGB image input units 39R, 39G, and 39B is fluctuated from the initial value only by the adjustment using the fixed parameter, the correct white balance adjustment cannot be performed, and the inspection accuracy is reduced. It becomes a deteriorating factor.
[0008]
The present invention has been made to solve such conventional drawbacks, and uses a color line sensor camera that can sufficiently cope with the white balance that varies over time with a simple configuration and always enables correct adjustment. An object is to provide a defect inspection apparatus.
[0009]
[Means for Solving the Problems]
According to the present invention, when an inspection object on which characters are printed in color and a timing marker is imaged, R, G, and B image signals of the inspection object are input to the RGB image as the timing marker is detected. A defect inspecting apparatus in which the color conversion unit inputs the R, G, B image signals in which these image signals are adjusted by the white balance parameter, and performs color conversion to display on the display unit.
Conveying means for conveying the inspection object,
The said is positioned adjacent to the conveying means, and the reference member surface is divided white region and the black region is provided,
A camera having three types of line sensors for imaging the inspection object and the reference member, and imaging red, green, and blue images;
From the image signal from the camera, only the first image signals of red, green, and blue in the white region of the reference member and the second image signals of red, green, and blue in the black region are extracted. A reference plate image input unit
The red, green, and blue white balance parameters are calculated based on an average level of luminance of the white region of the first image signal, an average level of luminance of the black region of the second image signal, and a target level. A white balance parameter calculation processing unit;
An RGB density conversion unit that sends an output signal obtained by density-converting the image signal of the inspection object from the RGB image input unit to the color conversion unit based on the white balance parameter. The gist.
[0010]
In the present invention, when the conveying means conveys the inspection object, a camera having three types of line sensors images the reference member arranged adjacent to the conveying means and the inspection object.
[0011]
Then, the white balance parameter calculation processing unit calculates the white balance parameter of each color of the white area and the black area of the reference member imaged by the camera.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a diagram showing a main configuration of an embodiment of the present invention. The same parts as those in FIG. 3 are denoted by the same reference numerals, and redundant description is omitted in principle. The sheet 32 to be inspected with the characters 34 printed on the surface thereof in color is conveyed in the direction of the arrow. The characters 34 with color printing are picked up for inspection by the three-plate CCD line sensor camera 33. Each color component signal of the captured image is input from the RGB CCDs 38R, 38G, and 38B to the RGB image input units 39R, 39G, and 39B, and corrected by the RGB density conversion units 41R, 41G, and 41B that perform white balance adjustment. Is done. Each RGB signal that has undergone the white balance correction processing is converted into an H (hue), S (saturation), and I (lightness) system by the color conversion unit 43. The subsequent inspection processing is performed using this HSI signal.
[0013]
Next, white balance adjustment, which is a feature of the present invention, will be described. First, the reference plate 11 is arranged in the vicinity of the conveyance path of the sheet 32 to be inspected and within the imaging range of the three-plate CCD line sensor camera 33. By arranging in this way, the reference plate 11 can be used together with the image to be inspected when the inspection range is picked up by the camera even if the inspection target article is sequentially subjected to inspection processing and conveyed. Images of 11 white and black areas 11w and 11b are also taken. That is, at the same time that the inspection target image is captured over time, the image of the reference plate 11 is also captured over time.
[0014]
The reference plate 11 is divided into two regions from the center up and down, one of which is formed as a white region 11w and the other as a black region 11b. Both the white region 11w and the black region 11b are optically standard white and black. That is, the white region 11w is formed so that the red (R), green (G), and blue (B) components are all equal in significance value, and the black region 11b is the component of each RGB color. It is formed to be zero.
[0015]
Now, when the inspection range is imaged by the three-plate CCD line sensor camera 33, the image of the white region 11w and the black region 11b of the reference plate 11 together with the image of the sheet 32 including the color characters 34 printed on the surface of the sheet 32. Is also imaged. As a result, the image of the character 34 and the image of the reference plate 11 are output together from the CCDs 38R, 38G, and 38B corresponding to RGB. Therefore, the image signal of the character 34 to be inspected is input to the RGB image input units 39R, 39G, and 39B, and the image signal of the reference plate 11 is input to the reference plate image input unit 12. The images of the white area 11w and black area 11b of the reference plate 11 input to the reference plate image input unit 12 are output for each RGB component by the RGB CCDs 38R, 38G, and 38B.
[0016]
The data of the reference plate image input unit 12 is input to the white balance parameter calculation processing unit 13, and the white balance parameters calculated for each RGB are respectively input to the R density conversion unit 41R, the G density conversion unit 41G, and the B density conversion unit 41B. Input and white balance correction is performed. As described above, when the inspection target is imaged within a predetermined range for the purpose of inspection, the reference plate is also imaged at the same time, and the images of the white area 11w and the black area 11b can be extracted, and a white balance parameter is created using the result. As a result, even when the inspection environment changes, the optimum white balance can be corrected in real time.
[0017]
Now, color signals of red (R), green (G), and blue (B) components are extracted from the black area 11b and the white area 11w of the reference plate 11 input to the reference plate image input unit 12, and the white balance parameter is extracted. The calculation processing unit 13 calculates a white balance parameter for each color component. These white balance parameters are provided to the RGB density conversion units 41R, 41G, and 41B. FIG. 2 schematically shows a white balance parameter calculation operation, and both are described by taking red (R) as an example.
[0018]
FIG. 2A shows a case where the black signal Rb and the white signal Rw from the reference plate 11 are in an optimum state where no white balance correction is required. Input signal (R
[Outside 1]

Because the case of is an ideal black, it means that the appropriate black has been captured [Outside 2]

This indicates that no white balance correction is required for the output signal R ′. The average value of the black signal and the white signal means an average luminance in the rectangular area in which a certain rectangular area is set in the black area 11b and the white area 11w of the reference plate.
[0019]
[Outside 3]

It will deviate from. For this reason, the black signal is corrected to be lowered by 30 and the white signal is corrected to be raised by 55, so that it can be taken in as an appropriate color signal. Here, only red (R) has been described as an example, but by performing the same correction process for other green (G) and blue (B), each color signal can be captured as an appropriate signal.
[0020]
The above relationship is expressed by mathematical formulas as follows. Where R is the input signal,
[Outside 4]

The average level, Lt, is the target level (the white area of the reference plate is converted to this label). In addition, although the above is an example of red (R), in the formula, green is G and blue is B, and each has the same meaning as described above.
[0021]
[Expression 1]

[Expression 2]

[Equation 3]

Appropriate white balance adjustment is performed by substituting the numerical value for each color into each of the above equations and setting the result in the corresponding density converter.
[0022]
【The invention's effect】
According to the present invention, a reference plate composed of a white area and a black area is provided in the vicinity of a position where the inspection object is mounted, and simultaneously, the white area and the black area of the reference plate are imaged. The white balance can be adjusted based on the signal from the camera, and thus the white balance can be adjusted in real time.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram showing an embodiment of the present invention.
FIG. 2 is a graph for explaining the operation of the present invention.
FIG. 3 is a schematic configuration diagram showing a conventional technique.
[Explanation of symbols]
11 Reference plate 11w White region 11b Black region 12 Reference plate image input unit 13 White balance parameter calculation processing unit 32 Sheet 33 Three-plate CCD line sensor camera 38R R-CCD
38G G-CCD
38B B-CCD
39R R image input unit 39GG G image input unit 39B B image input unit 41R R density conversion unit 41GG G density conversion unit 41B B density conversion unit 43 Color conversion unit

Claims (2)

When an image of an inspection object on which characters are printed in color and a timing marker is imaged, the RGB image input means inputs R, G, B image signals of the inspection object along with the detection of the timing marker. A defect inspection apparatus in which the color conversion unit inputs the R, G, and B image signals in which these image signals are adjusted with white balance parameters, performs color conversion, and displays them on the display unit,
Conveying means for conveying the inspection object,
The said is positioned adjacent to the conveying means, and the reference member surface is divided white region and the black region is provided,
A camera having three types of line sensors for imaging the inspection object and the reference member, and imaging red, green, and blue images;
From the image signal from the camera, only the first image signals of red, green, and blue in the white region of the reference member and the second image signals of red, green, and blue in the black region are extracted. A reference plate image input unit
The red, green, and blue white balance parameters are calculated based on an average level of luminance of the white region of the first image signal, an average level of luminance of the black region of the second image signal, and a target level. A white balance parameter calculation processing unit;
An RGB density conversion unit that sends an output signal obtained by density-converting the image signal of the inspection object from the RGB image input means based on the white balance parameter to the color conversion unit. A feature defect inspection device. RGB density converter
2. The defect inspection according to claim 1, wherein the density of each image data to which the parameter signal obtained by the white balance parameter calculation processing unit and the image data to be inspected from the three types of sensors are input is converted. apparatus.

Images