JPH0414067B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus for inspecting patterns on printed matter.

As shown in FIG. 1, a conventional print pattern inspection device scans a print 1 in the width direction with a line sensor camera 2 and extracts image density information based on a reflection device, while a rotary encoder attached to a cylinder 3 of a conveyance system scans the print 1 in the width direction. By synchronizing the scanning of the camera 2 with the output of 4, the sample pattern 5 (FIG. 2) is divided into a large number of small pixels in a matrix. The density of each of these pixels is sequentially compared with the corresponding small pixels of the sample pattern 6 (FIG. 2), which has been similarly divided and recorded in the memory, to detect the presence or absence of a defect.

Here, when inspecting the pattern of printed matter, it is necessary to set a certain tolerance value in order to avoid erroneously determining genuine paper as waste paper. This is necessary because there is some variation in density etc. even on regular paper, the printing position is not always at the same position in a strict sense, and the presence of noise is unavoidable. An appropriate tolerance value is given to the difference in reflected light quantity gradation with the sample pixel.

This tolerance value differs for each pixel depending on the picture, and an example is as follows. In other words, for pixels near the edges of a picture, the above-mentioned gradation difference becomes large due to positional deviation in the print conveyance system, so a larger tolerance value is given (Japanese Patent Application No. 179,599/1986).
No.). In addition, since the gradation difference due to positional deviation is small in flat parts of the pattern,
It is sufficient to give the pixels in such a portion a tolerance value that does not cause the correct paper to be erroneously determined to be waste paper due to variations in the original paper or noise. Therefore, a permissible number of defective pixels is set for the total number of defective pixels exceeding the predetermined permissible value, and paper is judged as waste or good paper based on whether the number of defective pixels exceeds the permissible number. It's on. In other words, this method inspects the pattern based on the combination of the allowable number of defects and the allowable tone difference value given to each pixel.

The relationship between this allowable number of defects and the allowable gradation difference judgment level (%) given to each pixel can be expressed by the characteristic shown in Figure 3, and this characteristic shows the following. There is. That is, in order to detect spotty and high-contrast stains caused by ink splatter among various defects, for example, it is necessary to increase the permissible judgment level value and reduce the permissible number of defects. Furthermore, in order to detect subtle color tone defects over the entire image caused by poor color matching, it is necessary to set the gradation difference tolerance value small and the defect tolerance number large. The shaded area above the characteristic curve in Figure 3 is the setting range that can be inspected, and the non-shaded area below the curve is the setting range that incorrectly judges normal paper as waste paper. It is necessary to set an area closer to the lower right, and in cases of poor color matching, it is necessary to set an area closer to the upper left.

However, since conventional printed matter pattern inspection equipment inspects using only one combination of gradation difference tolerance and defect tolerance, the characteristics of the pattern defects that can be detected are fixed, resulting in a wide variety of pattern defects. unable to respond. As a countermeasure for this, it is necessary to install a separate defect extraction device, which increases the size of the device and
This will lead to complications.

The present invention has been made in consideration of the above points, and
A pattern defect inspection method for printed matter that can deal with various pattern defects by setting a plurality of combinations of a tolerance value for the gradation difference between a sample pixel and a sample pixel and a tolerance value for the number of defective pixels exceeding this tolerance value. and equipment.

An embodiment of the present invention will be described below with reference to FIGS. 4 to 8.

FIG. 4 shows the overall configuration of a printed matter pattern inspection apparatus according to the present invention. In this device, a pattern on a printed matter 1 is scanned by a camera 2 to extract image information. At this time, the rotation amount of the conveyance system cylinder 3 is taken out by the rotary encoder 4 and given to the synchronization circuit 7. Synchronous circuit 7
provides its output to the camera 2, the A/D converter 8 and the address circuit 9, and performs scan synchronization of the camera 2;
Synchronization of A/D conversion of image information taken out by the camera 2 and address assignment of the memory 11 are performed. A switch SW inserted on the output side of the A/D converter 8 is used to switch between giving the output of the A/D converter 8 to the memory 11 or to the judgment circuit 10. When the camera 2 is extracting image information from the sample print, turn the switch SW toward the dashed line side to provide the image information (hereinafter referred to as specimen data) to the memory 11, and when the camera 2 is extracting image information from the sample print, turn the switch SW The image information (hereinafter referred to as sample data) is provided to the judgment circuit 10 by moving it toward the solid line side shown in the drawing, and is compared with the sample data read out from the memory 11 at the same time to detect pattern defects. In the figure, among the two input signals of the determination circuit 10, S indicates sample data and T indicates sample data, respectively.

FIG. 5 shows the internal structure of the determination circuit 10 in the apparatus shown in FIG. 4, which in this case is composed of n determination elements JE ₁ , JE _{2 .} . . JE _o and an OR gate. Each judgment element JE ₁ , JE ₂
... JE _o is given sample data S and sample data T, and its outputs J ₁ , J ₂ ...J _o are given to OR gate OR, and output TJ of OR gate OR becomes a comprehensive judgment signal.

Each determination element has a gradation difference tolerance and a tolerance number of defects set, and the values differ for each element. For example, JE ₁ has a large gradation difference tolerance and a small defect tolerance, and is designed to detect spot stains.
JE _o , on the other hand, has a small tone difference tolerance and a large defect tolerance, allowing it to detect subtle tonal differences over the entire image. And JE ₂ ~
JE _o-1 is considered to be an intermediate setting between JE ₁ and JE _o . As a result, JE ₁ and JE _o detect defects with extreme properties of picture defects, and JE ₂ to _{JE o-1} detect defects with properties intermediate between them.

When any one of these determination elements JE ₁ to JE _o produces an output indicating that the paper is waste paper, a waste paper signal is output from the OR gate OR.

Figure 6 shows the decision element in the circuit of Figure 5.
This shows the internal structure of JE. In this circuit, sample data S and sample data T are applied to an absolute difference circuit 22, and |S-T| is taken out and applied to the B input of a comparator 23. On the other hand, the allowable value E is given to the A input of the comparator from the allowable value generating circuit 21 based on the sample data S.

There are various ways to give this allowable value E. For example, a different tolerance value may be fixedly given to each pixel. Here, a value obtained by multiplying the gradation of a sample pixel by a fixed ratio (0 to 1) is set as an allowable value to be given to the sample pixel corresponding to that sample pixel. As a result, the allowable value generating circuit 11 multiplies the sequentially inputted sample data S by a fixed ratio (0 to 1) to generate a gradation difference allowable value E.

The comparator 23 determines whether the absolute value of the difference |S-T| is larger than the gradation difference tolerance E (|S-T|>
E) is determined and when it is large, the defect signal is sent to counter 2.
Give to 4. The counter 24 counts defective signals,
This count value represents the number of defective pixels. This count value is compared with the allowable number of defects from the latch 26 by the comparator 25, and it is determined whether the number of defective pixels exceeds the allowable number of defects, and when the number exceeds the allowable number of defects, a waste paper signal J is output.

FIG. 7 shows the internal structure of the gradation difference tolerance generating circuit 11 in the circuit of FIG. 6, and this circuit is composed of the same number of buffers B as the number of bits of the input signal. In this case, since the sample data S is 8 bits, there are 8 buffers, of which 2
One uses ground as input and its output as MSB,
Assume that the next 6 bits are the upper 6 bits of sample data S, including the MSB, taken out via buffer B. The LSB2 bits of the sample data S are not connected. As a result, the 8-bit sample data S is
The sample data S is shifted by 2 bits in the LSB direction.
A gradation difference tolerance value E of 12.5% is obtained. By changing the number of bits shifted in the same way
Values such as 50%, 25%, 6.25%, 3.125%, etc. are obtained.
To obtain values other than this, a multiplier such as TTL can be used.

FIG. 8 shows the internal configuration of the absolute difference circuit 22 in the circuit shown in FIG. 6, which includes an adder 31, an exclusive OR circuit 32, an adder 33, and an inverter 34 as main components. Since this circuit is well known, a detailed explanation will be omitted, but it obtains sample data S and sample data T and outputs the absolute value of the difference |S-T|.

As described above, the present invention sets a plurality of combinations of a tolerance value for gradation difference between a sample pixel and a sample pixel and a tolerance value for the number of defective pixels exceeding this tolerance value, and detects defects by any of these combinations. Since it is determined that there is a pattern defect when a pattern is detected, it is possible to reliably detect a wide variety of defects, unlike conventional methods in which the nature of the defects that can be detected is fixed. can be significantly improved. Furthermore, the device configuration is simple.

[Brief explanation of drawings]

Fig. 1 is a diagram showing the configuration of the image information detection section of the pattern inspection device that is the object of the present invention, Fig. 2 is a schematic explanatory diagram of sample data for pattern inspection and the recording method of the specimen data, and Fig. 3 is a characteristic diagram showing the relationship between the allowable judgment level for defect detection and the allowable number of defects,
FIG. 4 is a block diagram showing the configuration of the pattern inspection device according to the present invention, FIG. 5 is a circuit diagram showing the internal configuration of the determination circuit in the circuit of FIG. 4, and FIG. 6 is the determination in the circuit of FIG. 7 is a circuit diagram showing the internal configuration of the element, FIG. 7 is a circuit diagram showing the internal configuration of the gradation difference tolerance setting circuit in the circuit of FIG. 6, and FIG. 8 is a circuit diagram of the absolute difference value circuit in the circuit of FIG. 6. FIG. 3 is a circuit diagram showing the internal configuration. 1... Printed matter, 2... Camera, 3... Cylinder, 4... Rotary encoder, 5, 6... Memory, JE... Judgment element, OR... Or gate, B... Buffer, S... Sample data, T...
Sample data.

Claims (1)

[Scope of Claims] 1. A method for detecting defects in a pattern by decomposing a pattern of a sample print into a pixel matrix and sequentially comparing the density gradation of each pixel with a corresponding pixel in the pattern of the sample print, comprising: Multiple combinations of a tolerance value for the gradation difference between pixels and a tolerance value for the number of defective pixels exceeding the tolerance value are set, and if there is a defect that corresponds to any of these combinations, it is determined that there is a pattern defect. A method for inspecting patterns of printed matter, characterized by: 2. A device that generates a synchronization signal according to the conveyance operation of the printed material; a camera that scans the pattern of the printed material based on the synchronized signal to extract image information; and a memory into which the image information extracted from the sample printed material is read by the camera. and a plurality of determination elements each having a tolerance value for a gradation difference between the sample pixel and the sample pixel and a plurality of combinations of the tolerance value, and the determination element is configured to perform the determination on the image information of the sample print from the camera. A printed matter defect inspection device includes a determination circuit that outputs a defect signal when any of the elements is determined to be defective.