JP5303503B2 - Image inspection apparatus, printing apparatus, and image inspection method - Google Patents

Description

  The present invention relates to an image inspection apparatus that inspects an image printed on a print medium.

  In recent years, high-speed printing such as direct mail has been performed using an ink jet printing apparatus. In such a printing apparatus, individual information (so-called variable information) is printed on a portion corresponding to each page of printing paper that is roll paper. In an image printed by an ink jet printing apparatus (hereinafter referred to as “printed image”), ink omission due to nozzle clogging may occur due to adhesion of dust or a lump of ink. Therefore, an inspection device that detects such a defect is mounted on the printing apparatus.

  In defect inspection of printed matter in which the same image is printed repeatedly, a sample image is prepared by capturing a print image that does not have a defect, and an image obtained by capturing a print image to be inspected (hereinafter referred to as “inspection image”). ) And the sample image are compared for defect inspection. By acquiring the sample image and the inspection image with a common camera, it is possible to make the brightness and resolution the same by removing the influence of uneven printing and fluctuations in the amount of light, etc., and defect inspection is relatively easy. Is called. However, when variable information is printed, a sample image cannot be prepared, and it is necessary to directly compare an original image (data) used for printing with an inspection image. In this case, since the resolution and brightness are different between the original image and the inspection image, defect inspection becomes difficult.

  Regarding defect inspection of printed matter including variable information, the inkjet printing unit inspection apparatus disclosed in Patent Document 1 integrates pixel values in a predetermined direction with respect to a variable data target image, which is an image to be inspected, and inspects X Axial projection data is generated. Further, the inspection reference X-axis projection data is generated by integrating the pixel values in the same direction with respect to the variable data portion inspection reference image which is the reference image. In the inspection target X-axis projection data, the positional deviation and the magnification error with respect to the inspection reference X-axis projection data are corrected, and binarized with a threshold value close to 0 to generate inspection target zero extraction data. Also in the inspection reference X-axis projection data, a portion having a value of 0 is extracted to generate inspection reference zero extraction data. Ink missing data is generated by removing the zero portion common to the inspection reference zero extraction data from the inspection target zero extraction data. Ink jet printing unit inspection devices detect ink loss based on ink loss data.

  In the ink jet printer disclosed in Patent Document 2, an identification pattern is printed between a plurality of print images on rotary paper, and a print image is confirmed by checking whether there is ink dripping or ink loss in the identification pattern. The printing failure is confirmed.

JP 2003-94627 A Japanese Patent Application Laid-Open No. 10-100412

  By the way, in the method shown in Patent Document 1, in the case of color printing, even if ink loss of one color occurs in a mixed color portion, there is ink of another color, so the value obtained by integrating pixel values is not necessarily large. This is not always the case, and there are cases where ink loss cannot be detected. In addition, in a portion where the ink density is low, even if no ink omission occurs, the value obtained by integrating the pixel values of the inspection image easily increases due to the impact accuracy of the ink, and a defect may be erroneously detected. is there. Furthermore, in the ink jet method, variation occurs even in normal printing due to individual differences in printing paper and nozzles, and therefore, it is not possible to easily determine a threshold value for an integral value for determining that ink is missing. .

  Inkjet printing devices cause ink loss to occur during printing or to be eliminated. As shown in Patent Document 2, even if the presence or absence of defects occurring on the identification pattern is confirmed, printing can be performed. The presence or absence of defects in the image cannot be properly determined. Furthermore, since it is necessary to print an identification pattern having the same width as the print image in the width direction of the rotary paper, a space for printing the identification pattern is required on the top and bottom of each page. However, for example, in the case of printing paper on which symbols are printed across the perforations between pages, a space for printing the identification pattern cannot be secured.

  The present invention has been made in view of the above problems, and an object thereof is to improve the inspection accuracy of image inspection.

The invention according to claim 1 is an image inspection apparatus that inspects an image printed on a print medium, and that is generated in accordance with an imaging unit that captures the printed image and obtains an inspection image, and data used for printing A first integration unit that obtains a reference integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated values by integrating the values of pixels arranged in a predetermined printing direction in the reference image that has been determined; and the inspection A second integration unit that obtains an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in the printing direction in the image; the reference integrated value distribution; and the inspection In at least one of the integrated value distributions, a sensitivity correction unit that performs sensitivity correction to increase or decrease each integrated value by a predetermined process, and the reference integrated value after the sensitivity correction And a comparing unit for comparing the fabric with the inspection accumulated value distribution, wherein the reference image and the inspection image is a color image, the first integration unit, the second integration unit, the sensitivity correction unit and the comparison unit However, the sensitivity correction includes a correction that causes a maximum value filter to act on the reference integrated value distribution, and the sensitivity correction unit sets a predetermined value in the reference integrated value distribution. When the exceeding integrated value is changed to the upper limit value, the comparing unit excludes the integrated value exceeding the predetermined value from the comparison target .

The invention according to claim 2 is an image inspection apparatus for inspecting an image printed on a print medium, and is generated according to an image pickup unit that picks up the printed image and obtains an inspection image, and data used for printing A first integration unit that obtains a reference integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated values by integrating the values of pixels arranged in a predetermined printing direction in the reference image that has been determined; and the inspection A second integration unit that obtains an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in the printing direction in the image; the reference integrated value distribution; and the inspection In at least one of the integrated value distributions, a sensitivity correction unit that performs sensitivity correction to increase or decrease each integrated value by a predetermined process, and the reference integrated value after the sensitivity correction A comparison unit that compares the cloth with the inspection integrated value distribution, wherein the reference image and the inspection image are color images, and the first integration unit, the second integration unit, the sensitivity correction unit, and the comparison unit Is executed for each color component, and the sensitivity correction includes a correction for applying a maximum value filter to the reference integrated value distribution, and the first integrating unit and the second integrating unit are predetermined values. The process is executed for an area having a lower brightness.

A third aspect of the present invention is the image inspection apparatus according to the second aspect, wherein the sensitivity correction unit changes an integrated value exceeding a predetermined value in the reference integrated value distribution to an upper limit value. Thereby, the integrated value exceeding the predetermined value is excluded from the comparison target in the comparison unit.

  A fourth aspect of the present invention is the image inspection apparatus according to any one of the first to third aspects, wherein the sensitivity correction multiplies or divides each integrated value by a constant, or adds or subtracts a constant. Including corrections to be made.

The invention according to claim 5 is the image inspection apparatus according to any one of claims 1 to 4 , further comprising a differentiation processing unit for differentiating the reference integrated value distribution and the inspection integrated value distribution, and the comparison The unit compares the reference integrated value distribution after differentiation with the inspection integrated value distribution.

A sixth aspect of the present invention is the image inspection apparatus according to any one of the first to fifth aspects, wherein the reference image and the inspection image are divided into a plurality of blocks, and the first image is divided into corresponding blocks. One accumulation unit, the second accumulation unit, the sensitivity correction unit, and the comparison unit execute processing.

Invention of claim 7, a printing apparatus, a printing mechanism for printing on a print medium by free edition claims 1 to inspect the image printed in parallel to the printing by the printing mechanism 6 Or an image inspection apparatus according to any one of the above.

The invention according to claim 8 is the printing apparatus according to claim 7 , wherein the printing mechanism performs printing by ejecting fine droplets of ink from a head portion, and the head portion is configured to perform printing. Each position of the print area on the print medium is passed only once.

The invention according to claim 9 is an image inspection method for inspecting an image printed on a printing medium, wherein a) a reference image is prepared according to data used for printing, and b) a printed image is captured. And c) a reference integrated value which is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in a predetermined printing direction in the reference image. A step of obtaining a distribution; d) a step of obtaining an inspection integrated value distribution which is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in the printing direction in the inspection image; ) Performing a sensitivity correction to increase or decrease each integrated value by a predetermined process in at least one of the reference integrated value distribution and the inspection integrated value distribution; and f) the sensitivity With the reference accumulated value distribution after the correction and the step of comparing the inspection accumulated value distribution, wherein the reference image and the inspection image is a color image, said c) through the f) step is performed for each color component The step e) includes a correction for applying a maximum value filter to the reference integrated value distribution. In the step e), an integrated value exceeding a predetermined value in the reference integrated value distribution becomes an upper limit value. As a result of the change, the integrated value exceeding the predetermined value is excluded from the comparison target in the step f) .
The invention according to claim 10 is an image inspection method for inspecting an image printed on a print medium, wherein a) a reference image is prepared according to data used for printing, and b) a printed image is captured. And c) a reference integrated value which is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in a predetermined printing direction in the reference image. A step of obtaining a distribution; d) a step of obtaining an inspection integrated value distribution which is a distribution in a direction perpendicular to the printing direction by integrating the values of pixels arranged in the printing direction in the inspection image; ) Performing sensitivity correction to increase or decrease each integrated value by a predetermined process in at least one of the reference integrated value distribution and the inspection integrated value distribution; and f) the above A step of comparing the reference integrated value distribution and the inspection integrated value distribution after degree correction, wherein the reference image and the inspection image are color images, and the steps c) to f) are performed for each color component. Executed, and the step e) includes a correction that causes a maximum value filter to act on the reference integrated value distribution. In the step c) and the step d), the region having a lightness lower than a predetermined value is used. The process is executed.

According to the present invention, inspection accuracy can be improved by comparing integrated value distributions after sensitivity correction. Further, it is possible to it is possible to improve the inspection accuracy for the color image, to reduce the effect of deviations in comparing the two images. In the invention of claim 4, the two integrated value distributions can be brought close by a simple calculation, and as a result, the comparison becomes easy.

In the present invention, furthermore, by excluding light portion that inspection accuracy is degraded, Ru can reduce erroneous detection of a defect. In the inventions of claims 1 and 9 , the bright part can be easily excluded from the comparison target. In the invention of claim 5 , even if there is a difference in brightness between the reference image and the inspection image, the inspection can be performed accurately. According to the sixth aspect of the invention, it is possible to reduce the influence of the brightness difference in the inspection image on the defect accuracy.

It is a figure which shows a printing apparatus. It is a bottom view which shows the discharge part of a printing mechanism. It is a figure which shows an image reading apparatus. It is a figure which shows an image inspection part. It is a figure which shows the flow of a defect inspection. It is a figure which illustrates a reference image. It is a figure which illustrates an inspection image. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows inspection integrated value distribution. It is a figure which shows a reference | standard integrated value distribution and a test | inspection integrated value distribution. It is a figure which shows difference distribution. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows reference | standard integrated value distribution. It is a figure which shows the integrated value distribution of a sample image. It is a figure which shows the curve produced | generated based on the integrated value distribution of a sample image. It is a figure which illustrates the inspection image in other examples of processing. It is a figure which illustrates a reference image. It is a figure which shows the other example of a calculating part. It is a figure which shows standard differential value distribution. It is a figure which shows inspection differential value distribution. It is a figure which shows standard differential value distribution. It is a figure which shows a reference | standard differential value distribution and test | inspection differential value distribution.

  FIG. 1 is a perspective view showing an appearance of a printing apparatus 10 according to an embodiment of the present invention. The printing apparatus 10 includes a printing mechanism 11 that performs printing on a printing sheet 9 that is a web without a plate, a conveyance mechanism 12 that conveys the printing sheet 9, an image inspection apparatus 13 that inspects an image, and a control unit that controls these. 14. The printing apparatus 10 performs printing of an image including variable information, so-called variable printing, and automatically inspects an image on the printing paper 9 (hereinafter referred to as “printed image”). The control unit 14 rasterizes data of an image to be printed (hereinafter referred to as “original image”), thereby generating data used for printing (hereinafter referred to as “original image data”). Thus, the control unit 14 also serves as an image data generation unit.

  The transport mechanism 12 transports the printing paper 9 relative to the printing mechanism 11 in the (−Y) direction in FIG. Note that the X direction, the Y direction, and the Z direction in FIG. 1 are perpendicular to each other, and the Z direction corresponds to the vertical direction. In the transport mechanism 12, a plurality of rollers 121 that are each long in the X direction in FIG. 1 are arranged in the Y direction. On the (+ Y) side of the plurality of rollers 121, a supply unit 122 that holds a roll of the printing paper 9 before printing and feeds the printing paper 9 from the roll toward the printing mechanism 11 is provided. On the (−Y) side of the plurality of rollers 121, a winding unit 123 that winds and holds a portion on which the printing paper 9 has been printed in a roll shape is provided. In the following description, when the printing paper 9 is simply referred to, it means the printing paper 9 being transported (that is, the printing paper 9 on the plurality of rollers 121).

  The printing mechanism 11 is disposed above the transport mechanism 12 and is fixed to a frame 151 provided on the base 15 so as to straddle the transport mechanism 12. FIG. 2 is a bottom view of one ejection unit 111 of the printing mechanism 11. In FIG. 2, the moving direction (that is, the (−Y) direction) of the printing paper 9 with respect to the ejection unit 111 is shown upward. Actually, the printing mechanism 11 is provided with a plurality of ejection units that eject inks of cyan (C), magenta (M), yellow (Y), and black (K), respectively. Arranged in the Y direction. A plurality of heads 112 are arranged in the ejection unit 111 in a direction perpendicular to the moving direction of the printing paper 9 and parallel to the printing surface of the printing paper 9 (the X direction in FIGS. 1 and 2 and corresponding to the width of the printing paper 9). Therefore, a plurality of discharge ports 1121 are arranged at a constant pitch in the width direction on the bottom surface of each head portion 112. The

  The head unit 112 is provided with a piezoelectric element corresponding to each ejection port 1121, and a minute droplet of ink is ejected toward the printing paper 9 from each ejection port 1121 by driving the piezoelectric element. The distance between the two head portions 112 adjacent to each other in the width direction is adjusted with high accuracy, and all the discharge ports 1121 included in the discharge portion 111 extend over the entire width of the print area on the printing paper 9 in the width direction. They are lined up at a constant pitch. In the printing apparatus 10, printing is performed at high speed only by passing each position of the printing area of the printing paper 9 once under the ejection unit 111 (so-called one-pass). In FIG. 2, only five discharge ports 1121 are shown in each head unit 112, but in reality, a large number of discharge ports 1121 are arranged. The head unit 112 may be of a type that ejects fine droplets by heating ink.

  In the printing apparatus 10, in parallel with the movement of the printing paper 9 in the (−Y) direction by the transport mechanism 12, the control unit 14 controls ink ejection from the head unit 112 of each color according to the original image data to be printed. By doing so, a color print image for one page is formed on a portion of the print paper 9 corresponding to one page of the printed matter. In the following description, the (+ Y) direction in FIG. 1, which is the direction in which an image is printed by the printing mechanism 11, that is, the direction opposite to the moving direction of the printing paper 9 is referred to as a “printing direction”.

  The image inspection device 13 includes an image reading device 131 that reads a printed image of the printed paper 9 after printing (that is, a portion where printing of the printing paper 9 is finished), and an image inspection unit 132 that inspects the printed image. The image reading device 131 is located on the downstream side of the printing mechanism 11.

  FIG. 3 is a view of the image reading device 131 viewed from the width direction of the printing paper 9, and shows a simplified internal configuration of the image reading device 131. The image reading device 131 includes a first line illumination unit 21a, a second line illumination unit 21b located on the (+ Y side) than the first line illumination unit 21a, and the first line illumination unit 21a and the second line illumination unit 21b. The imaging part 22 located above is provided. The first line illumination unit 21a includes a plurality of light emitting diodes 211 and a lens unit 212 arranged in the X direction. Light from the plurality of light emitting diodes 211 is made uniform in the X direction, and is printed along the optical axis J1. 9 is irradiated linearly. Similarly, the second line illumination unit 21b includes a plurality of light emitting diodes 211 and a lens unit 212, and the light from the plurality of light emitting diodes 211 is made uniform in the X direction, and the first line illumination unit along the optical axis J2. The same position as the position irradiated by 21a is irradiated linearly. The imaging unit 22 images a linear region included in the region irradiated with light by the first line illumination unit 21a and the second line illumination unit 21b.

  The imaging unit 22 includes a color line camera 221, a first mirror 223, and a second mirror 224. The first mirror 223 has a strip shape that is long in the X direction, and is disposed below the color line camera 221. The second mirror 224 also has a strip shape that is long in the X direction, and is located on the (−Y) side of the color line camera 221. When imaging is performed by the imaging unit 22, the light from the printing paper 9 is incident on the second mirror 224 along the optical axis J <b> 3, is reflected by the second mirror 224, and travels to the first mirror 223. Led. The light reflected by the first mirror 223 enters the color line camera 221.

  FIG. 4 is a block diagram showing a functional configuration of the image inspection unit 132. In the present embodiment, the image inspection unit 132 is realized by a computer and a dedicated electric circuit. Of course, the entire image inspection unit 132 may be realized by a computer or a dedicated electric circuit. The image inspection unit 132 includes a calculation unit 31 that performs various calculation processes, a correction parameter setting unit 32 that sets correction parameters to be described later, a storage unit 33 that stores image data, inspection results, and the like, and a result output that outputs the inspection results The unit 34 is provided. In FIG. 4, storage devices such as a memory and a hard disk drive in the image inspection unit 132 are collectively shown as the storage unit 33. The calculation unit 31 includes a first integration unit 311, a second integration unit 312, a sensitivity correction unit 313, and a comparison unit 314. These functions will be described later.

  Next, an operation in which the printing apparatus 10 inspects a printed image while printing an image on the printing paper 9 will be described. FIG. 5 is a diagram illustrating a flow of inspection of a print image. First, the original image data to be printed is prepared by the control unit 14 shown in FIG. Next, in a RIP (Raster Image Processor) unit in the control unit 14, raster format original image data corresponding to the CMYK color space is generated from the vector format original image data and transmitted to the printing mechanism 11.

  In the printing apparatus 10, the printing paper 9 is continuously moved in the (−Y) direction by the transport mechanism 12, and the ejection of ink of each color is controlled according to the original image data by the printing mechanism 11. A printed image is formed on top. Then, a print image immediately after formation is picked up by the image reading device 131 located in the forward direction of the printing paper 9, that is, downstream of the printing mechanism 11, and a color image to be inspected (hereinafter referred to as "inspection image"). Are sequentially acquired (step S11). The inspection image is stored as inspection image data 62 in the storage unit 33 shown in FIG.

  On the other hand, in parallel with the printing operation, the original image data is sent from the RIP unit 141 of the control unit 14 to the image inspecting unit 132 and converted from CMYK color space to RGB color space data to indicate the reference image ( Hereinafter, it is stored and prepared as “reference image data 61”) (step S12).

  FIG. 6 is a diagram illustrating a reference image, and FIG. 7 is a diagram illustrating an inspection image. In FIG. 6, the downward direction of the reference image 41 in FIG. 6 corresponds to the print direction of the print image. The same applies to FIG. 7 and the following similar drawings.

  4 performs alignment of the reference image 41 and the inspection image 42 shown in FIGS. 6 and 7 with respect to the Y direction that is the printing direction and the X direction that is perpendicular to the printing direction (step). S13). In the first integration unit 311, the values of pixels arranged in the printing direction in the reference image 41 are integrated, and a distribution in the X direction of the integrated value is obtained as shown in FIG. 8 (step S <b> 14). Hereinafter, this distribution is referred to as “reference integrated value distribution 510”. Actually, a reference integrated value distribution and an inspection integrated value distribution, which will be described later, are acquired for each of the RGB color components. In the following description, only one color component will be described.

  In the reference image 41, the pixel value increases in a portion where no design exists, that is, in a portion corresponding to the background (white) of the printing paper 9. In the design portion, when the brightness of the color component to be calculated is low, the value of the pixel of the color component becomes small. Therefore, in the pixel column indicated by the broken line at the position 411 in the X direction in FIG. 6 (hereinafter, simply referred to as “pixel column”), there are many symbol portions, and as shown in FIG. The integrated value at the position 411 corresponding to the pixel column in the distribution 510 is minimized. As shown in FIG. 9, the sensitivity correction unit 313 performs later-described sensitivity correction for increasing or decreasing each integrated value in the reference integrated value distribution 510 indicated by a thin solid line (step S15), and sensitivity correction denoted by reference numeral 51 is performed. A later reference integrated value distribution is obtained.

  In parallel with the generation of the reference integrated value distribution 51, the values of the pixels arranged in the printing direction in the inspection image 42 shown in FIG. 7 are integrated by the second integration unit 312. As shown in FIG. A distribution in the X direction (hereinafter referred to as “inspection integrated value distribution 52”) is obtained (step S16). In the inspection image 42 shown in FIG. 7, ink loss due to ink ejection failure occurs in the pixel row at the position 421 indicated by the broken line, and the integrated value increases. As shown in FIG. 10, in the inspection integrated value distribution 52, a peak appears at a position 421 corresponding to the pixel column where the defect exists.

  FIG. 11 is a conceptual diagram showing the reference integrated value distribution 51 and the inspection integrated value distribution 52 after sensitivity correction in an overlapping manner. The comparison unit 314 subtracts the reference integrated value from the inspection integrated value at each position in the X direction (step S17). FIG. 12 is a diagram illustrating a difference distribution between the inspection integrated value and the reference integrated value. Hereinafter, this distribution is referred to as “difference distribution 53”. In the comparison unit 314, the portion where the difference is a negative value is changed to zero. As shown in FIG. 11, since only the portion where the defect exists in the inspection integrated value distribution 52 is located above the reference integrated value distribution 51, there is a defect in the difference distribution 53 as shown in FIG. 12. A positive value appears at the position where the defect occurs, thereby identifying the position of the defect.

  By the way, cyan, magenta and yellow inks in the printed image have the most influence on the inspection integrated value distribution corresponding to the complementary colors red, green and blue, respectively. In the defect inspection, for example, if there is a positive value in the difference distribution regarding red at a position where cyan and magenta are applied in the printed image, it is estimated that ink missing has occurred in cyan that absorbs red.

  Inspection result data 64 indicating the inspection result is stored in the storage unit 33 and is output to the operator via the result output unit 34 such as a display unit. Note that the inspection on the inspection image 42 may be performed every several lines instead of in units of pages. In the printing apparatus 10, print images for many pages are continuously formed on the printing paper 9. In parallel with the printing operation, the image inspection apparatus 13 continuously inspects defects in the printed image of each page immediately after formation. Done.

  Next, sensitivity correction by the sensitivity correction unit 313 will be described. In the sensitivity correction, first, in the reference integrated value distribution 510 illustrated in FIG. 8, as shown in FIG. 13, the integrated value equal to or higher than a certain threshold T is the upper limit, that is, only the pixels corresponding to the background portion of the printing paper 9. The accumulated value is obtained when accumulated in the printing direction. At the position changed to the upper limit value, the inspection integrated value distribution 52 in FIG. 10 is necessarily located below the reference integrated value distribution 511. Thereby, in the comparison part 314, the integrated value of the reference integrated value distribution 511 exceeding the threshold T is substantially excluded from the comparison target.

  If the integrated value is large, the increase in the integrated value is limited even if ink loss occurs at that position. Therefore, in a range in the X direction where the integrated value is large, ink loss is hardly detected or erroneous detection is likely to occur. This is a region where it is difficult to visually determine ink loss. Therefore, in the image inspection apparatus 13, by performing a process of extracting a dark part having a small integrated value in the reference integrated value distribution, the bright part can be easily excluded from the comparison target and defect detection errors can be reduced.

  Next, a filtering process is performed on the reference integrated value distribution 511. In the filtering process, by applying a maximum value filter to the reference integrated value distribution 511, each integrated value becomes maximum in a predetermined range including the integrated value (for example, a range of 5 pixels in the (± X) direction). Is replaced with another integrated value. FIG. 14 is a diagram showing the reference integrated value distribution 512 after the maximum value filter is applied, and the reference integrated value distribution before the maximum value filter is applied is indicated by a broken line. In FIG. 15 and FIG. 16 described later, the reference integrated value distribution after the previous process is indicated by a broken line. 14 to 16 are diagrams showing an outline of the processing, and do not show an accurate processing result.

  In a portion where the integrated value of the reference integrated value distribution 512 changes in the X direction, the integrated value is greatly increased by the maximum value filter. The part corresponds to the vicinity of the edge of the symbol portion in the reference image 41 (see FIG. 6). As a result, the integrated value of the inspection integrated value distribution 52 is larger than the integrated value of the reference integrated value distribution 512 in the vicinity of the edge due to incomplete alignment between the reference image 41 and the inspection image 42. Is prevented. In other words, the sensitivity of defect detection by the comparison unit 314 is weakened in the vicinity of the edge, and the influence of the shift between both images is reduced. As a result, erroneous detection of defects is further prevented.

  Further, the sensitivity correction unit 313 performs so-called gain processing in which the integrated value of the reference integrated value distribution 512 in FIG. 14 is multiplied by a positive constant. When the constant to be multiplied is larger than 1, the integrated value increases as the integrated value increases, and for example, a reference integrated value distribution 513 shown in FIG. 15 is obtained. Thereby, it is possible to further weaken the sensitivity in a large integrated value range while weakening the detection sensitivity in the comparison unit 314 in the entire X direction. When the value multiplied by the gain processing is a positive constant less than 1, the detection sensitivity is enhanced as a whole, and the sensitivity is further enhanced in the range where the integrated value is large. In the gain process, the integrated value may be divided by a predetermined positive constant.

  Next, so-called offset processing is performed in which a predetermined positive constant is added to or subtracted from the integrated value of the reference integrated value distribution 513. When a positive constant is added, the entire integrated value increases uniformly as in the reference integrated value distribution 514 shown in FIG. 16, so that the detection sensitivity in the comparison unit 314 is uniformly weakened as a whole. When a positive constant is subtracted from the integrated value (or a negative constant is added), the detection sensitivity is uniformly enhanced as a whole. By performing the gain process and the offset process, the reference integrated value distribution 514 and the inspection integrated value distribution 52 can be brought close to each other by a simple calculation, and as a result, appropriate comparison between them is facilitated.

  As described above, the sensitivity correction processing by the calculation unit 31 has been described. However, since the generation process is different between the reference image 41 and the inspection image 42, ink loss is easily detected by simply comparing the integrated value distributions as they are. On the other hand, in the image inspection apparatus 13, sensitivity correction is performed on the reference integrated value distribution by the image inspection unit 132, so that a difference in brightness and contrast between the reference image 41 and the inspection image 42, a positional deviation, and the like. Thus, the defect is detected from the inspection image 42 with high accuracy.

  Next, an example of a method for determining constants used for gain processing and offset processing, that is, values of correction parameters will be described. First, a print image that has been confirmed to have no defects is taken a plurality of times to obtain a picked-up image (hereinafter referred to as “sample image”), and an integrated value in the print direction is obtained for the plurality of sample images. It is done. FIG. 17 is a diagram illustrating three integrated value distributions 541 to 543 obtained by integrating pixel values of three sample images. Next, the correction parameter setting unit 32 in FIG. 4 selects the maximum of the integrated values of the three integrated value distributions 541 to 543 at each position in the X direction in FIG. 17, and as shown in FIG. A curve 54 connecting the largest integrated values is generated.

  Further, a reference image is generated from the print image data used to generate the sample image, and a reference integrated value distribution is acquired. Then, the integrated value of the reference integrated value distribution is multiplied or added to the integrated value and the positive real number so that the reference integrated value distribution approximates or slightly exceeds the curve 54. Or a combination with a positive real number to be subtracted from the integrated value is determined. The two positive real numbers are stored in the storage unit 33 as part of the correction parameter 65 in the gain process and the offset process, as shown in FIG.

  By using a plurality of sample images, it is possible to determine the gain processing and offset processing constants while reducing the influence of variations in the printed image due to the deflection of the printing paper 9 and the ink ejection accuracy.

  Note that the size of the maximum value filter, which is a part of the correction parameter 65, is determined according to the maximum amount of deviation assumed to remain between both images after the alignment of the reference image 41 and the inspection image 42, for example. The

  Although the configuration and operation of the printing apparatus 10 have been described above, the image inspection apparatus 13 of the printing apparatus 10 improves the inspection accuracy by comparing the reference integrated value distribution 51 and the inspection integrated value distribution 52 after sensitivity correction. be able to. The arithmetic unit 31 acquires the reference integrated value distribution and the inspection integrated value distribution for each of the RGB color components and performs defect inspection, thereby improving the inspection accuracy for the color image.

  The ink loss inspection using the integrated value is particularly suitable for a one-pass printing apparatus in which ink loss greatly affects print quality.

  As described above, when the integrated value in the reference integrated value distribution is large, even if ink loss occurs at that position, it is difficult to detect ink loss, or erroneous detection is likely to occur. Therefore, when the pixel values are integrated for each color component by the first integration unit 311 and the second integration unit 312, even if a value greater than a certain value in the reference image 41 and the inspection image 42 is excluded from the integration target. Good. As a result, in the subsequent processing, the processing is executed only for the region where the lightness is lower than the certain value in each color component. As a result, defect detection omission and false detection are suppressed.

  Further, in the image inspection unit 132, the brightness of the inspection image 42 changes due to changes in the sensitivity of the imaging unit 22 and the brightness of illumination over time, lens contamination, and the like. Therefore, constants used for offset processing and gain processing may be updated in real time. In this case, for example, the constant is determined so that the average value of the reference integrated value distribution is approximately the same as the average value of the inspection integrated value distribution.

  Next, another operation example of the image inspection unit 132 will be described. FIG. 19 is a diagram illustrating an inspection image acquired by the image inspection unit 132. In the image inspection unit 132, the inspection image 42a is inspected for each divided block. The configuration of the printing apparatus and other operations of the image inspection unit 132 are the same as those in FIGS. 1, 4, and 5.

  When a defect inspection is performed on a print image, first, the print image is picked up by the image reading device 131 in FIG. 1, and the inspection images are sequentially acquired. Further, the image inspection unit 132 generates the reference image data 61 of FIG. 4 based on the original image data. Next, the inspection image 42a is divided into a plurality of blocks having a predetermined number of pixels in the X direction and the Y direction. When a block surrounded by a broken line in FIG. 19 (hereinafter referred to as “target block 422”) is selected as an inspection target, the target block 422 and the reference image 41a shown in FIG. Alignment with the corresponding range of is performed.

  Specifically, a range in the reference image 41a larger than the block of interest 422 (hereinafter referred to as “search range 412”) is set, and the block of interest 422 is moved in the X and Y directions while moving in the block of interest 422. The sum of absolute values of differences between the values of the respective pixels and the values of the corresponding pixels in the search range 412 (hereinafter referred to as “difference total value”) is obtained. Then, the position of the block of interest 422 having the smallest difference value within the search range 412 is obtained as the position after alignment. The root mean square of the difference between the pixel value of the block of interest and the value of the overlapping pixel in the search range (that is, the norm of the pixel value difference) is obtained. May be determined as the position of

  In the image inspection unit 132, the same processing as that after step S14 in FIG. 5 is performed in the range of the reference image 41a that overlaps the block of interest 422 after alignment. By this process, the defect inspection process is executed by substantially dividing both the reference image 41a and the inspection image 42a into blocks. Specifically, a reference integrated value distribution in a range where the target block 422 overlaps in the reference image 41a is acquired, sensitivity correction is performed, and the reference integrated value distribution and the test integrated value distribution of the target block 422 are compared to determine whether ink has run out. Presence / absence is determined. When the inspection for one block is completed, the other blocks are sequentially inspected in the same manner, and the entire inspection image 42a is inspected. It should be noted that the defect inspection need not be performed in a block where no symbol exists.

  In the above operation example, by performing defect inspection by the calculation unit 31 for each block, it is possible to accurately detect ink loss in the dark block. Therefore, if there is even one dark block in the printing direction, the defect can be detected with high accuracy in the column where the block exists. Further, by dividing the inspection image 42a into a plurality of blocks, alignment can be performed with high accuracy. It should be noted that the defect inspection using the divided blocks of the reference image 41a and the inspection image 42a may be substantially performed by dividing the reference image 41a and aligning it with the inspection image 42a.

  FIG. 21 is a diagram illustrating another example of the calculation unit 31 of the image inspection unit 132. The computing unit 31 is provided with a differentiation processing unit 315 that performs a differentiation process on the reference integrated value distribution and the inspection integrated value distribution. Other configurations of the image inspection unit 132 are the same as those in FIG.

  When defect inspection is performed on a print image, first, the image inspection unit 132 sequentially acquires inspection images (data thereof) and generates a reference image (data thereof). Hereinafter, the reference image 41 in FIG. 6 and the inspection image 42 in FIG. 7 will be described as examples. Next, alignment between the inspection image 42 and the reference image 41 is performed. A reference integration value distribution 510 illustrated in FIG. 8 is obtained by the first integration unit 311, and gain processing is performed on the reference integration value distribution 510 by the sensitivity correction unit 313. Further, the differential processing unit 315 differentiates the integrated value with respect to the X direction, and as shown in FIG. 22, the distribution of the differential value of the reference integrated value distribution 510 is acquired. Hereinafter, the distribution is referred to as “reference differential value distribution 515”, and each differential value included in the reference differential value distribution 515 is referred to as “reference differential value”.

  In parallel with the generation of the reference differential value distribution 515, the second integration unit 312 obtains the inspection integration value distribution 52 similar to that in FIG. In the differential processing unit 315, the integrated value of the inspection integrated value distribution 52 is differentiated with respect to the X direction, and the distribution of the differential value of the inspection integrated value distribution 52 is acquired as shown in FIG. Hereinafter, the distribution is referred to as “inspection differential value distribution 521”, and each differential value included in the inspection differential value distribution 521 is referred to as “inspection differential value”. At the position where ink loss has occurred, the inspection integrated value distribution 52 changes sharply as shown in FIG. 10, so that a peak appears in the inspection differential value distribution 521.

  Next, an offset process is performed on the reference differential value distribution 515 so that the average value of the inspection differential value is the same as the average value of the reference differential value in FIG. 22 or the average value of the reference differential value is slightly increased. Gain processing is performed. In FIG. 24, the reference differential value distribution 515 before correction is indicated by a thin solid line, and the reference differential value distribution 515a after correction is indicated by a thick solid line. FIG. 25 is a conceptual diagram showing the inspection differential value distribution 521 and the corrected reference differential value distribution 515a in an overlapping manner. The comparison unit 314 subtracts the reference differential value distribution 515a from the inspection differential value distribution 521. Thereby, a positive value appears only at the position in the X direction where the defect exists, and the position of the defect is specified.

  In the image inspection apparatus 13 including the calculation unit 31 illustrated in FIG. 12, even if a contrast difference remains between the reference image 41 and the inspection image 42, the differentiated reference integrated value distribution and the inspection integrated value distribution are obtained. By comparing, defect inspection can be performed accurately.

  Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications can be made. In the above embodiment, in the sensitivity correction, dark part extraction processing, filtering processing, gain processing, and offset processing are performed, but it is not always necessary to perform all of them, and at least one is performed, thereby improving inspection accuracy. be able to. In sensitivity correction, the above processing may be performed in various orders. Sensitivity correction may be performed on the inspection integrated value distribution, or may be performed on both the reference integrated value distribution and the inspection integrated value distribution.

  In the above embodiment, the comparison unit 314 of the calculation unit 31 may determine the presence / absence of a defect by comparing the difference distribution obtained by subtracting the reference integrated value distribution from the inspection integrated value distribution and the threshold value. Further, a defect may be detected by dividing each integrated value of the inspection integrated value distribution by a corresponding integrated value of the reference integrated value distribution. In this case, it is determined that a defect exists at a position where a value obtained by dividing the integrated value of the reference integrated value distribution from the integrated value of the inspection integrated value distribution exceeds a predetermined threshold. In the comparison unit 314, the reference integrated value distribution and the inspection integrated value distribution may be compared by a method other than subtraction and division.

  The image inspection apparatus 13 may be used for other printing apparatuses that perform printing without plate such as an electrophotographic system in addition to the ink jet printing apparatus. A plate-like member may be used.

9 Printing paper 10 Printing device 11 Printing mechanism 13 Image inspection device 22 Imaging unit 41, 41a Reference image 42, 42a Inspection image 51, 510-514 Reference integration value distribution 52 Inspection integration value distribution 112 Head unit 311 First integration unit 312 First 2 Accumulation unit 313 Sensitivity correction unit 314 Comparison unit 315 Differentiation processing unit 422 Attention block 515, 515a Reference differential value distribution 521 Inspection differential value distribution

Claims (10)

An image inspection apparatus for inspecting an image printed on a print medium,
An imaging unit that captures a printed image to obtain an inspection image;
A reference integrated value distribution, which is a distribution in a direction perpendicular to the printing direction, is obtained by integrating the values of pixels arranged in a predetermined printing direction in a reference image generated according to data used for printing. 1 integration unit,
A second integrating unit that obtains an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of pixels arranged in the printing direction in the inspection image;
In at least one of the reference integrated value distribution and the inspection integrated value distribution, a sensitivity correction unit that performs sensitivity correction to increase or decrease each integrated value by a predetermined process;
A comparison unit for comparing the reference integrated value distribution and the inspection integrated value distribution after the sensitivity correction;
Equipped with a,
The reference image and the inspection image are color images, and the first integration unit, the second integration unit, the sensitivity correction unit, and the comparison unit execute processing for each color component,
The sensitivity correction includes a correction that causes a maximum value filter to act on the reference integrated value distribution,
When the sensitivity correction unit changes the integrated value exceeding the predetermined value in the reference integrated value distribution to the upper limit value, the comparing unit causes the integrated value exceeding the predetermined value to be compared from the comparison target. An image inspection apparatus that is excluded . An image inspection apparatus for inspecting an image printed on a print medium,
An imaging unit that captures a printed image to obtain an inspection image;
A reference integrated value distribution, which is a distribution in a direction perpendicular to the printing direction, is obtained by integrating the values of pixels arranged in a predetermined printing direction in a reference image generated according to data used for printing. 1 integration unit,
A second integrating unit that obtains an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of pixels arranged in the printing direction in the inspection image;
In at least one of the reference integrated value distribution and the inspection integrated value distribution, a sensitivity correction unit that performs sensitivity correction to increase or decrease each integrated value by a predetermined process;
A comparison unit for comparing the reference integrated value distribution and the inspection integrated value distribution after the sensitivity correction;
Equipped with a,
The reference image and the inspection image are color images, and the first integration unit, the second integration unit, the sensitivity correction unit, and the comparison unit execute processing for each color component,
The sensitivity correction includes a correction that causes a maximum value filter to act on the reference integrated value distribution,
The image inspection apparatus, wherein the first integration unit and the second integration unit execute processing on an area having a lightness lower than a predetermined value . The image inspection apparatus according to claim 2 ,
When the sensitivity correction unit changes the integrated value exceeding the predetermined value in the reference integrated value distribution to the upper limit value, the comparing unit causes the integrated value exceeding the predetermined value to be compared from the comparison target. An image inspection apparatus that is excluded. The image inspection apparatus according to any one of claims 1 to 3,
2. The image inspection apparatus according to claim 1, wherein the sensitivity correction includes a correction of multiplying or dividing each integrated value by a constant, or adding or subtracting a constant. The image inspection apparatus according to any one of claims 1 to 4 ,
A differential processing unit for differentiating the reference integrated value distribution and the inspection integrated value distribution;
The comparison unit compares the reference integrated value distribution after differentiation and the inspection integrated value distribution. An image inspection apparatus according to any one of claims 1 to 5 ,
The reference image and the inspection image are divided into a plurality of blocks, and the first integration unit, the second integration unit, the sensitivity correction unit, and the comparison unit execute processing for each corresponding block. Image inspection device. A printing device,
A printing mechanism for printing on a printing medium without printing;
And an image inspection apparatus according to any one of claims 1 to 6 for inspecting the images printed in parallel to the printing by the printing mechanism,
A printing apparatus comprising: The printing apparatus according to claim 7 , wherein
The printing mechanism performs printing by ejecting micro droplets of ink from the head unit,
The printing apparatus, wherein the head section passes through each position of a printing area on a printing medium only once during printing. An image inspection method for inspecting an image printed on a print medium,
a) preparing a reference image according to data used for printing;
b) capturing a printed image to obtain an inspection image;
c) obtaining a reference integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of pixels arranged in a predetermined printing direction in the reference image;
d) obtaining an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of the pixels arranged in the printing direction in the inspection image;
e) performing sensitivity correction to increase or decrease each integrated value by a predetermined process in at least one of the reference integrated value distribution and the inspection integrated value distribution;
f) comparing the reference integrated value distribution with the inspection integrated value distribution after the sensitivity correction;
Equipped with a,
The reference image and the inspection image are color images, and the steps c) to f) are performed for each color component,
The step e) includes a correction for applying a maximum value filter to the reference integrated value distribution,
In step e), an integrated value exceeding a predetermined value in the reference integrated value distribution is changed to an upper limit value, so that an integrated value exceeding the predetermined value is compared in step f). An image inspection method characterized by being excluded from a target . An image inspection method for inspecting an image printed on a print medium,
a) preparing a reference image according to data used for printing;
b) capturing a printed image to obtain an inspection image;
c) obtaining a reference integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of pixels arranged in a predetermined printing direction in the reference image;
d) obtaining an inspection integrated value distribution that is a distribution in a direction perpendicular to the printing direction of the integrated value by integrating the values of the pixels arranged in the printing direction in the inspection image;
e) performing sensitivity correction to increase or decrease each integrated value by a predetermined process in at least one of the reference integrated value distribution and the inspection integrated value distribution;
f) comparing the reference integrated value distribution with the inspection integrated value distribution after the sensitivity correction;
Equipped with a,
The reference image and the inspection image are color images, and the steps c) to f) are performed for each color component,
The step e) includes a correction for applying a maximum value filter to the reference integrated value distribution,
An image inspection method , wherein in the step c) and the step d), processing is performed on an area having a lightness lower than a predetermined value .

Images