JPH0410861B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a printed matter inspection method for in-line comparing the state of a printed matter being printed with a standard state in a printing press to detect abnormalities in the printed matter.

Conventionally, the mainstream method for inspecting printed matter has been offline and relying on human vision. This stems from the fact that each piece of printed matter has a different pattern, and because inspection items for printed matter have been thought to involve subtle differences that require reliance on human vision. On the other hand, in response to the desire to evaluate printed matter while it is being printed, attempts have been made to use strobe lighting that is synchronized with the printing speed and to use mirrors that rotate synchronously at high speed to judge printed matter that is being printed as a still image. It was done. However, these methods were not systems that could be called inspection machines in that they relied on human vision for inspection.

In addition, attempts have been made to print color patches at the same time as the patterns on the printed matter and inspect the color patches, thereby allowing the inspection of the printed matter to be performed on behalf of the user. However, with this method, printing problems (oil dripping,
If dirt (such as dirt) occurs on the pattern, it will be missed, and the inspection machine cannot be said to be functioning satisfactorily.

On the other hand, a system has recently been proposed that inspects printed matter in-line using a line sensor, as seen in the "Printed Material Inspection Apparatus" disclosed in Japanese Patent Application No. 57-220515. By using this system, the pattern itself of printed matter can be automatically inspected in-line, so it does not have the above-mentioned drawbacks and can be expected to have excellent effects as an inspection machine.

However, the above invention is not without its problems; for example, during printing, the pattern on the printing paper may be out of phase in the vertical and horizontal directions due to tension fluctuations, changes in drying temperature, and other factors. In-line inspection is hampered by the problem that it is difficult to distinguish an abnormal signal caused by a phase shift of an inspection image from a true abnormal signal.

For this reason, Japanese Patent Application No. 172777/1983 has proposed a method for correcting synchronization defects in inspection of running printed matter, and by utilizing this invention, the phase shift in the machine direction in a rotary printing press can be minimized. This makes it possible to perform inspections with reduced pressure.

However, there are cases in which simply utilizing the above invention is not sufficient.

In other words, there is a danger that a slight phase shift will be detected as a large signal difference at the edge of a photograph, etc. in a printed image, that is, at the boundary between the white part of the paper and the high density part of the ink. .

For example, if we assume that we are performing 1m/m pixel detection on the boundary line between a white background (0% dot) and 100% dot, then with a phase shift of 0.1m/m,
An abnormal signal of 10% level will be generated. In this case, it can be said that it is extremely difficult to distinguish between a true abnormal signal due to a printing abnormality and a false abnormal signal due to a phase shift.

In order to solve this problem, Japanese Patent Application Laid-Open No. 58-81165
As seen in ``Inspection Method for Printed Materials'' by No. 1, a method has been proposed in which the tolerance of the inspection signal for the edge area is relaxed compared to other areas, but in the above proposal, if a defect occurs in the edge area, the defect is There is a high possibility that something will be overlooked, and this is insufficient in practice.

The present invention solves these problems,
By utilizing the present invention, it is an object of the present invention to enable more accurate inspection by distinguishing between a false abnormal signal at the edge of a picture caused by a phase shift and a true abnormal signal due to an abnormality in the printed matter.

Hereinafter, the present invention will be described in detail based on some examples.

FIG. 1 is a schematic diagram showing the configuration of a printed matter inspection apparatus according to the present invention. Although FIG. 1 shows an example of attachment to a rotary printing press, there is no problem even if the apparatus is attached to a sheet-fed printing press. In Fig. 1, a printing paper 3 fed from a roll-shaped paper roll 2 is printed in four colors (yellow, red, indigo, and black) on both sides in a printing section 1, and then transferred to a dryer and a folding machine (see Fig. 1). (not shown). In order to inspect the printing condition after four colors have been printed on both sides, the printed matter inspection device detects pattern information on the printing paper using the detection section 4 while timing the inspection sampling with the rotary encoder 5 provided in the printing section 1. The line sensor inputs the information to the processing circuit 6, and a judgment operation is made as to whether or not the pattern information matches the reference information.

As a result, if it is determined that the printing condition is abnormal, it becomes possible to take measures such as alarms, markings, rejects, etc.

However, as mentioned above, in a printing machine, the pattern on the printing paper 3 is affected by tension fluctuations, drying temperature fluctuations, etc., and it is impossible to completely eliminate the phase shift using only the timing pulses from the rotary encoder 5. It is.

For this reason, as shown in FIG. 2, in boundary areas where the density difference is large, false abnormal signals due to phase shifts are likely to occur. That is, FIGS. 2A and 2B show the level of the reference signal and the test signal when a portion including a boundary where the density difference of the pattern is large is detected at the same timing based on the rotary encoder 5 when a phase shift occurs. As shown in the figure, if there is a shift e in the signal acquisition timing based on the phase shift and a comparison test is performed based on the difference between the reference signal and the test signal, as shown in Figure 2 C. As a result, the difference signal exceeds the allowable range S, and therefore, even though the printed material in this area is normal, it is erroneously determined to be abnormal.

In the present invention, the CCD provided in the detection section 4
The optical detection means such as the above is adjusted so that the light receiving surface of the photoelectric conversion element is out of focus (so as to be out of focus) from the state where it is originally focused. This results in
The reference signal and test signal shown in Figure 2 A and B are as follows:
As shown in Figures 3A and 3B, they gradually change (the signal becomes dull) at the boundary, so if you take the difference, the pulses caused by the phase shift are optically blurred, as shown in Figure 3C. This can be suppressed to a much smaller level than the tolerance range S, which has the height and width associated with the amount, the density difference at the boundary, and the phase shift amount, and the influence of the phase shift can be eliminated.

However, if the image is simply defocused and optically blurred, true printing abnormalities, such as abnormalities with a small area and a difference in level, cannot be detected. That is, when the reference signal is a flat signal as shown in Figure 4A, in the original focused state,
As shown by the dotted line in Figure 4B, a printing abnormality that can be perceived as a small abnormal signal with a considerable level difference from the reference signal is different from a small abnormality with a gradual level difference as shown in the solid line because it is out of focus. It is detected as a signal (dull signal), and even the difference falls within the tolerance range S as shown in FIG. 4C, making it impossible to detect.

In order to eliminate such defects, the following processing is performed in the present invention.

In other words, a difference is calculated between the reference signal and the test signal detected when the focus is out of focus, the signals obtained by the difference calculation are added for a certain number of images, and the addition result is determined within an acceptable range. Performs processing to determine whether or not it is within the range.

Here, considering that the phase shift occurs randomly with respect to the pattern position, for example, a phase shift of one page in the scanning direction occurs in five consecutive prints,
Let us assume a case where a phase shift occurs in one sheet in the reverse scanning direction and phase shifts in the other three sheets are sufficiently small.

The reference signal at the boundary between the optically blurred low-density area and the high-density area is as shown in Figure 5 (a), and when the aforementioned phase shift occurs, each test signal is as shown in Figure 5 (b). As shown in the figure, a phase shift in the scanning direction is represented by 1, a phase shift in the reverse scanning direction is represented by m, and a sufficiently small phase shift is represented by n. By calculating the difference between each test signal for these five pictures and the reference signal shown in Figure 5A, and adding the results of this difference calculation, no signal exceeding the allowable range appears as shown in Figure 5C. do not have.

This is due to the fact that the phase shift does not tend to advance or lag, but instead occurs randomly, so the false abnormal signals caused by the phase shift cancel each other out, and the amount of phase shift decreases. Since the frequency of occurrence of large cases is low, even if the difference signals for consecutive pictures are added as described above, the result of the addition will not exceed the allowable range.

On the other hand, printing abnormalities that occur in small areas and have different levels, such as hits and ink splatters, tend to occur on multiple sheets in succession at the same position of the image. For example, as shown in Figure 6A, for a flat signal,
As shown in Figure 6 (b) for five consecutive pictures, one picture has a phase shift in the scanning direction, one picture has a phase shift in the reverse scanning direction, and three pictures have almost no phase shift. When a printing abnormality occurs in a small area with a level difference, the optical detection means is out of focus and is detected as a gradual signal with a small level difference, as shown in Figure 6 (b). Ru. Here, o, p, and q indicate detection signals in a state where there is almost no phase shift when a phase shift occurs in the scanning direction, when a phase shift occurs in the reverse scanning direction, and when there is almost no phase shift.

By calculating the difference between the inspection signal shown in Figure 6B and the reference signal shown in Figure 6B, and adding the values for 5 images, the level difference in the area where the printing abnormality has occurred is found as shown in Figure 6C. It is possible to extract a large abnormality signal, sufficiently exceeding the allowable range S, and to determine whether or not a printing abnormality has occurred.

In other words, an abnormal signal that did not exist in the reference signal is detected as a signal with a gradual and small level difference due to the optical halo, but the signal after the difference calculation is composed of components in the same direction. There is no change, and as a result, by adding up a plurality of pictures in time series, it is possible to compensate for the deficiencies caused by the blurring of the focus of the optical detection means.

In addition, the number of additions of the above-mentioned chronological patterns is
It is preferable to set the number of sheets to about 4 to 10 in consideration of the nature of continuous occurrence of printing failures, the characteristics of phase shift, etc. After the addition, a comparison is made with the allowable range S, and the addition result is refreshed each time.

In addition, in order to dull the detection signal, we used an optical halo phenomenon, but there are other methods that can be considered, such as using an electrical method. Without processing, it is difficult to obtain a signal in only one direction (XO to Y), which has the disadvantage of complicating the configuration, so it is preferable to utilize the optical halo phenomenon as in the present invention.

In addition to the difference calculation in the process described above, as seen in the "Inspection Method for Printed Materials" in Japanese Patent Application No. 58-172778, the signal after the difference calculation and the signal after this difference calculation are delayed by several pixels. It is also possible to perform other processing, such as a process of further calculating a difference between the added signal (so-called secondary difference), before the addition calculation.

Next, an example of a processing circuit for realizing the printed matter inspection method of the present invention will be explained based on FIG.
Here, an optical halo is used to dull the input signal, and an optical detection means (not shown) is adjusted and maintained in an out-of-focus state.

A detection unit 4 using an optical detection means such as a CCD line sensor detects the information of the picture printed on the printing paper using a sampling timing signal calculated by a sampling control circuit 15 based on pulses generated by a rotary encoder 5. Input is performed by.

Since the input image signal IS is an analog signal, it is converted into a digital signal via the A/D converter 7. When the printing operator presses the instruction button from the operation stand to set the current printed matter as the reference, the digitized input signal is stored in the reference memory 8, and is used as the reference when inspection is started from now on. Used repeatedly as a signal.

When the test is started, the input signal is transferred from the A/D converter 7 directly to the difference circuit 9. At this time, the memory control circuit 14 reads the reference signal in the reference memory 8 while specifying the address of the reference memory 8, and transfers it to the difference circuit 9 at the same time as the input signal, so that the difference circuit 9 outputs one test signal. −Difference signal DS of reference signal 1
is output.

Here, in comparing the reference signal and the test signal, the method is not limited to the difference method only, and the above-mentioned second-order difference method etc. may also be used. The circuit configuration described above can be applied, and other circuit configurations that can obtain similar results may also be used.

Subsequently, the difference signal DS is stored in the addition memory 10, and at the same time, the difference signal DS is stored in the addition memory 1 by the addition circuit 16.
The value within 0 is added and transferred to the comparison circuit 11, where it is checked by the CPU 12 whether the specified tolerance range is exceeded. When testing the first picture, the value in the addition memory 10 is zero. if,
If an abnormality is found in the first pattern, the comparison circuit 11
The error signal ES is transferred to the CPU 12, and the elimination means 13 such as alarm, marking, reject, etc.
can be used to deal with defective printed matter.

Here, if no abnormality is found in the first pattern, the memory control circuit 14 calls out the difference signal of the first pattern stored at the same address from the addition memory 10 for the difference signal of the next pattern. , both are added in the adder circuit 16, and the result is transferred to the comparator circuit 11,
Tests similar to those above are performed. Further, this addition result is stored in the addition memory 10. Thereafter, after the addition, storage and inspection are repeated up to a predetermined number of patterns, the contents of the addition memory 10 are refreshed and the above process is repeated.

Here, these additions, recordings, and erasures are controlled by the memory control circuit 14. The CPU 12 can instruct the memory control circuit 14 to add the pictures.

Note that the processing circuit shown in FIG. 7 is an embodiment for realizing the method of the present invention, and there is no problem in using other circuit configurations or software processing.

As described above, according to the printed matter inspection method according to the present invention, by utilizing the optical halo phenomenon, false abnormal signals caused by the phase shift at the boundary between the low-density part and the high-density part of the picture are detected. In addition to adding the difference signals for a certain number of consecutive pictures and determining whether or not the result is within the allowable range, even when there is a phase shift, there is still a small All printing abnormalities, including printing abnormalities that appear as signals with large level differences in area, can be reliably detected.

[Brief explanation of drawings]

The drawings show one embodiment of the present invention, and FIG. 1 is a schematic diagram of a printed matter inspection apparatus to which the present invention is applied, and FIG. A model diagram of the false abnormal signal that occurs. Figure 3 is a model diagram showing the mitigation of phase shift by using an optical halo. Figure 4 is a model diagram showing the influence of an optical halo on small-area defects. figure,
Figure 5 shows that by using addition memory,
A model diagram showing the effect of mitigating the generation of abnormal signals due to phase shift. Figure 6 is a model showing the effect of using addition memory to remove the effects of optical halo on abnormal signals caused by small-area defects. 7 are block diagrams showing one embodiment of a processing circuit for carrying out the print inspection method of the present invention. 3... Print paper, 4... Detection section, 5... Rotary encoder, 8... Reference memory, 10...
Addition memory, 11... Comparison circuit, 12...
CPU.

Claims (1)

[Claims] 1. Image pattern information of the printed material is captured pixel by pixel, and an inspection signal based on the detected pattern information of each pixel is compared with a reference signal of the corresponding pixel to inspect abnormalities occurring in the printed material. In the method of A printed matter inspection method characterized by adding the results for each corresponding pixel of consecutive printed matter for a certain number of sheets, and determining whether or not this added signal exists within a predetermined tolerance range.