JP3417178B2 - Circuit pattern inspection method and inspection apparatus - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inspection method and an inspection apparatus for detecting a defect in a circuit pattern of a printed wiring board, a mask film, a semiconductor integrated circuit, etc., and particularly to inspect by comparing pattern detection data with reference data. The present invention relates to an inspection method and an inspection device.

[0002]

2. Description of the Related Art Circuit patterns include printed circuit boards and semiconductor integrated circuits on which electric circuits are formed, and circuit patterns for forming these electric circuits. For example, a circuit pattern of a mask film, a circuit pattern of an exposed substrate, a circuit pattern obtained by screen-printing an etching resist ink, a copper circuit pattern after etching, and the like. The miniaturization of these circuit patterns has progressed, and defects such as pattern short circuits, disconnections, chips, protrusions, thickening, thinning, and residual patterns are likely to occur in the manufacturing process, and improvement in yield is a major issue.

Further, the electrical characteristics inspection of these finished products cannot be said to be sufficient inspection, and in order to guarantee the quality of the circuit, visual inspection or visual inspection by an automatic machine is usually performed in each manufacturing process. The visual appearance inspection has problems of nerve fatigue and oversight of defects, and a shift to automatic appearance inspection is desired due to the problem of labor saving. Particularly, when the circuit pattern width is 250 μm or less, the efficiency of visual inspection by a person is lowered, and the occurrence rate of overlooking is increased, which is a problem.

As a method of automatic visual inspection, a comparison method and a design rule method are mainly used. The comparison method includes comparison with a design pattern, comparison with a non-defective pattern, and the like, which requires a large memory capacity, but is a simple method. On the other hand, the design rule method can reduce the memory capacity, but in many cases, the circuit pattern to be inspected contains patterns that deviate from the general design rule, and it is often not possible to deal with the design rule method alone.

Therefore, recently, these methods are often combined. However, the above-mentioned comparison method and design rule method tend to cause excessive detection, require confirmation work (verification) after automatic inspection, and cannot be inlined, which results in low productivity and requires a person to confirm. was there. From this point of view, the future method is to create the reference data that is the standard for good products rather than the good product pattern, because of the simplicity of the system, the low cost of the memory, and the compatibility with patterns outside the design rules. The method of comparing and judging with is promising. However, even in this method, since inspection is performed in accordance with a fine circuit pattern, there remains a problem that pseudo defects that do not become defects in low-density portions, dust, and foreign matter are excessively detected. However, if the problem of excessive detection can be solved, the method of comparing and determining the inspection pattern data and the reference data will be far superior to other methods.

For example, in the case of a circuit pattern as shown in FIG. 2, in a fine pattern portion, even minute breaks, protrusions, disconnections, short circuits and pinholes indicated by a _{2 to} a ₆ may damage the circuit function and are detected as defects. There is a need to.
Therefore, since the reference data corresponding to a fine circuit pattern is set so that such a minute defect can be detected, in addition to the large hole a ₁ and the large remaining pattern a _{7 in} the low density circuit pattern portion, there is a problem in the circuit function. Fine cracks, pinholes, small residual patterns, etc. (b _{2 to} b ₅ )
Had been detected as a defect. Further, there is a problem that the small deviation b ₁ of the mounting hole H of the circuit board, small dust, foreign matter, surface flaw, and dirt are excessively detected as defects and the yield is deteriorated.

As a method for eliminating these excessive detections, n × m is applied to the image data in which the circuit pattern is detected.
There is a method of performing a smoothed image processing such as a weighted average processing or a median filter for each (for example, 3 × 3) pixel matrix. However, in such a smoothed image processing method, the defect detection performance for each pixel decreases, Further, since the processing is uniform regardless of the circuit parts, it is not possible to sufficiently discriminate and delete the excessive detection with high accuracy.

[0008]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and an object thereof is an inspection method for inspecting by comparing detected pattern data of a circuit pattern with reference data. In a device, there is provided a circuit pattern inspection method and a device thereof capable of detecting an original defect without deteriorating the defect detection performance of one pixel unit and finally detecting what can be regarded as a non-defect. .

[0009]

In order to achieve the above object, according to the invention of claim 1, the detected pattern data created by imaging the circuit pattern of the object to be inspected is compared with the reference data to detect the defect of the circuit pattern. In the method of inspecting the circuit pattern to be detected, the primary judgment process of detecting the defect correction by comparing the detection pattern data with the reference data in which the allowable range of the detection pattern data is set, and the primary judgment process is determined to be the defect correction. A secondary determination process for determining a defect / non-defect based on the detection pattern data of the pixel is set by setting a pixel group of a pixel which has a predetermined arrangement centered on the pixel.
It has a predetermined arrangement of the orbiting pixel group according to the size of the pixel group.
Multiple kinds of pixels are provided, and the primary revolving pixel until it is judged as non-defective
It is characterized in that the secondary determination is sequentially repeated from the group to the final orbiting pixel group .

[0010]

The invention of claim 2 is characterized in that, in the invention of claim 1 , the orbiting pixel group having the smallest size is a primary orbiting pixel group. According to the invention of claim 3, in the invention of claim 1 or 2 , in the secondary determination process,
It is characterized in that it is determined as non-defect if the variation width of the detection pattern data of the surrounding pixel group is within a predetermined uniformity.

According to a fourth aspect of the invention, in the invention of the third aspect, a plurality of types of predetermined uniformity are set, and the predetermined uniformity is switched and used according to the magnitude of the average value of the detection pattern data of the surrounding pixel groups. it is characterized in.

According to a fifth aspect of the invention, in the first or second aspect of the invention, the detection pattern data of the revolving pixel group is data of both the circuit portion and the non-circuit portion of the circuit pattern to be inspected in the secondary determination process. It is characterized in that it is judged as a defect by including. In the invention of claim 6 , claims 1 to
In any of the inventions of 5 to 5, the circuit portion, the non-circuit portion of the circuit pattern to be inspected based on the reference data of the revolving pixel group,
Alternatively, it is characterized in that it is discriminated whether it is a specific portion or a boundary portion thereof, and a predetermined arrangement of a plurality of kinds of surrounding pixel groups is selected and used.

In the invention of claim 7 , what is claimed in claims 1 to 6 is
In one of the inventions, it is determined from the reference data of the orbiting pixel group whether it is a circuit portion, a non-circuit portion, or a specific portion of the circuit pattern to be inspected, or a boundary portion thereof,
It is characterized in that a predetermined uniformity of a plurality of types of detection pattern data variation widths of the circulating pixel group is selected and used.
According to an eighth aspect of the present invention, in any one of the first to seventh aspects of the present invention, a plurality of pieces of area division data that divides an inspection area that is created in advance are used The present invention is characterized in that a predetermined arrangement of the peripheral pixel group for which the seed is set is selected and used.

In the invention of claim 9 , what is claimed in claims 1 to 7 is
In one of the inventions, using the area division data that divides the inspection area that is created in advance, the area division data of the pixel determined to be a defect candidate in the primary determination process is used to determine the detection pattern data of a plurality of types of the surrounding pixel group. Is selected and used .

[0016]

According to a tenth aspect of the present invention, the image pickup means for detecting the detection pattern data of the object to be inspected, the detection pattern data section for storing the detection pattern data output from the image pickup means, and the allowable range of the detection pattern data are set. A reference data creation unit that creates set reference data, a reference data unit that stores the reference data created by the reference data creation unit, and a primary determination that detects the defect candidate by comparing the detection pattern data with the reference data. And a secondary determination for determining a defect / non-defect based on detection pattern data of the peripheral pixel group, the peripheral pixel group having a predetermined arrangement centered on the pixel determined as the defect candidate by the primary determination unit. And a secondary determination means for detecting the detection pattern of the orbiting pixel group.
If the fluctuation range of the image data is within the predetermined uniformity, it is judged as non-defect.
It is a means for determining .

According to an eleventh aspect of the present invention, in the tenth aspect of the present invention, the secondary determining means uses the circuit data, the non-circuit portion of the circuit pattern to be inspected based on the reference data of the revolving pixel group,
Alternatively, it is a means for discriminating whether or not it is a specific portion or a boundary portion thereof, and selecting and using a predetermined arrangement of a plurality of orbital pixel groups set or a predetermined uniformity of a variation width of a plurality of set detection pattern data. Characterize.

According to the invention of claim 12 , claim 10 or 1 is provided.
In the first aspect of the present invention, an area division data section that creates and stores the area division data that divides the inspection area is provided, and the secondary determination means uses the area division data of the pixels determined to be defect candidates by the primary determination means, The present invention is characterized in that it is a means for selecting and using a predetermined arrangement of a plurality of kinds of circular pixel groups set or a predetermined uniformity of the fluctuation width of a plurality of set detection pattern data.

[0020]

Therefore, in the invention of claim 1 and the invention of claim 10 , in the defect correction of the primary judgment, the distance to the circuit edge portion on the circuit pattern and the size of the defect candidate can be judged together. It is possible to reduce unnecessary detection by eliminating unnecessary defect correction (pseudo defect). Further, for a large defect, it is possible to reduce the number of defective defective pixels inside and narrow down to only a part of the pixels, which allows efficient data storage and post-processing. Particularly, in the invention of claim 1, it is relatively large.
Even if it is a pseudo defect, it is not missing if it is separated from the edge of the circuit pattern.
This can result in a pitfall, which can reduce over detection.
The invention of claim 10 is simple and suitable for high-speed processing,
Enables efficient and high-speed inspection.

[0022]

According to the second aspect of the present invention, even if there are many small pseudo defects, since processing is performed from a small group of surrounding pixels, efficient and high-speed processing can be performed, and even for large defects, the number of defective defective pixels inside is reduced. High-speed processing can be efficiently performed by processing from a small group of circular pixels. According to the invention of claim 3 , simple and suitable for high-speed processing, and efficient and high-speed inspection can be performed.

According to the fourth aspect of the present invention, it is possible to more accurately determine the non-defect by switching the uniformity reference according to the size of the average value of the detection pattern data of the peripheral pixels for defect correction, and it is possible to detect excessive detection. Can be further reduced . Especially the detection pattern
According to the average size of the
By replacing the parts, the parts on the circuit pattern can be discriminated, and the homogeneity criterion suitable for each can be discriminated with high precision, and excessive detection can be further reduced.

According to the invention of claim 5 , the defect candidate near the circuit edge of the circuit pattern can be recognized by a simple algorithm, and highly reliable inspection can be performed. According to the sixth , seventh and eleventh aspects of the present invention, it is possible to accurately determine the defect occurrence site on the circuit pattern, and it is possible to further reduce excessive detection by making a determination according to each site.

According to the eighth , ninth and twelfth aspects of the present invention, since the area division data can be set, it is possible to judge the quality based on a separately determined judgment standard for a specific area of the circuit pattern, and to cope with exceptional situations. It is possible to reduce these excessive detections .

[0027]

[0028]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. (Embodiment) FIG. 1 shows an overall configuration of an embodiment of an inspection apparatus using the method of the present invention. In the figure, a table 3 is driven by an NC (numerical control) drive unit 2 to move in the arrow direction in the figure. And an image pickup camera 1 which is arranged above the table 3 and is composed of a line CCD camera for picking up an image of the circuit board 4 to be inspected placed on the table 3, and the circuit board 4 to be inspected which is imaged by the image pickup camera 1. Lamp 5 for illuminating the surface of the object with linear light in a direction orthogonal to the moving direction, and A / D for A / D converting the image signal of the imaging camera 1.
The conversion unit 6 and the illumination unit 7 that controls the illumination of the illumination lamp 5.
And the imaging camera 1, the illumination unit 7, the NC drive unit 2 and the A / D
The image pickup means I is constituted by the image pickup controller 8 which controls the operation of the converter 6. Here, the circuit board 4 to be inspected may be fixed and the side of the imaging camera 1 may be NC driven. Further, fine adjustment positioning means (electrical, mechanical and combinations thereof) in the X, Y and θ directions for alignment with the reference data are incorporated.

Further, the digital image signal from the A / D converter 6 is taken in under the control of the image pickup controller 8, and X, Y address values corresponding to the X, Y positions of each pixel of the image of the circuit board 4 to be inspected. A detection pattern data section 9 for storing detection pattern data consisting of luminance data, and a plurality of circuit patterns serving as a standard for inspection are picked up in advance, and a circuit is provided for each of the maximum value and the minimum value of the brightness value of each of these detection pattern data. A reference data creation unit 10 for creating an upper limit reference value and a lower limit reference value by adding an allowable value for each part of a pattern, and a reference data creation unit 10
The upper limit reference value and the lower limit reference value generated in the above are stored in the detection pattern data unit 9 and a reference data unit 11 that stores the same as the detection pattern data as reference data having X, Y address values corresponding to the X, Y positions. After the positions of both the reference data and the detection pattern data of the detected pattern data are aligned, the upper limit value stored in the reference data unit 11,
The primary determination unit 12 that sequentially detects the defect candidate of the circuit pattern of the circuit board 4 to be inspected by sequentially comparing it with the reference data including the lower limit for each pixel, and the pixel determined as the defect candidate by the primary determination unit 12 A circular pixel group having a predetermined arrangement centered on a pixel is set, the detection pattern data corresponding to the circular pixel group is read from the detection pattern data unit 9, and the maximum and minimum luminance values for each pixel of the circular pixel group are set. Find the average value,
If the average value is 50% or more or less than 50% of the brightness range, a reference value of uniformity is selected, and if (maximum value-minimum value)> uniformity reference value, it is determined as a defective pixel on the circuit pattern, and A secondary determination unit 13 that counts the number of pixels determined to be non-defective and the number of pixels determined to be the final defect for each orbiting pixel group, and outputs the secondary determination result, and each pixel of the entire captured image The image data processing means II is constituted by the area division data section 14 which stores the area division data for switching the contents of the secondary judgment processing corresponding to the X and Y positions. Here, the output 12a of the primary determination unit 12 indicates processing of defective candidate pixels in the primary determination, and the output 12b indicates processing of non-defective pixels in the primary determination.

Next, the above primary judgment, reference data, area division, etc. will be described. As shown in FIG. 2, the circuit pattern includes a circuit portion, a non-circuit portion, a specific portion such as a mounting hole H, and a boundary portion thereof. FIG. 2 illustrates a circuit pattern defect, a pseudo defect such as dust, a flaw, and a stain, and a micro defect that is a circuit pattern defect but does not cause a problem in circuit function as a final product.

The size criterion, which is a criterion for determining whether or not the defect is a pseudo defect, differs depending on the generation site in the circuit pattern. That is, the reference is strict in the fine circuit pattern portion, and the reference is loosened in the low density and rough portion. By loosening this standard, it is possible to reduce over-detection and perform efficient and high-speed inspection. In the example of the circuit pattern of FIG. 2, a _{1 to} a ₇ is a true defect and b _{1 to} b ₅ is a pseudo defect.

Here, the drawings on the right side and the lower part of the circuit pattern of FIG. 2 show projection diagrams of the pixel value levels, for example, the luminance values on the line Y and the line X of the circuit pattern, and in each drawing, the line Y, The upper limit reference value L ₁ of the reference data of the pixel located on X and the lower limit reference value L ₂ are overlaid and displayed. Then, the black portion of the profile exceeds the upper or lower reference value L ₁ or L ₂ and is detected as a defect candidate.

Then, the primary judgment unit 12 sequentially judges, for each pixel, the upper limit / lower limit reference value of the detection pattern data and the reference data, and in the case of FIG. 2, a _{1 to} a ₇ , b _{1 to} b.
₅ is determined as a defect candidate. The detection pattern data and the reference data are aligned in the X, Y, and θ directions by a method of correcting the positional deviation before the detection pattern data is imaged, a method of correcting the positional deviation during the imaging, and a position after image processing by image processing. There is a method of correcting the deviation, and it is necessary to perform accurate positioning by using these methods at the time of primary determination.

Further, in this example, the circuit portion is imaged brightly and the non-circuit portion is imaged darkly. On the contrary, the circuit portion is imaged darkly.
Even if the non-circuit part is brightly imaged, the same processing can be performed. In this way, if only the determination of the primary determination unit 12 is completed, b ₁
~ B ₅ is excessively detected as a defect. In the present invention,
The secondary determination unit 13 is provided to prevent excessive detection.

Next, the operation of the secondary judgment unit 13 will be described with reference to FIGS.
A description will be given based on FIG. First, as shown in the flowchart of FIG. 5, the secondary determination unit 13 sets the pixel (Pi, j) determined to be a defect candidate by the primary determination unit 12 on the detection pattern data corresponding to the X and Y positions of the circuit pattern as shown in FIG. As described above, a predetermined circle pixel group (filter F ₁ ) is set. As shown in FIG. 4A, the filter F1 is composed of a 3 × 3 pixel group (shown by diagonal lines) centered on the defective pixel Pi, j.

Next, the brightness data of each pixel of the set revolving pixel group is read, and the maximum value of the read brightness data is Z _max , the minimum value is Z _min , and the average value is Za. If the average value Za is 50% or more of the luminance range Z _R , the secondary determination unit 13 regards it as a circuit portion and sets the uniformity reference to C1.
If the average value Za is less than 50% of the brightness range Z _R , it is regarded as a non-circuit portion, and the uniformity reference is set to C2. And Z
_max −Z _min ≦ C (where C is C ≧ 1 if Za ≧ Z _R × 0.5, and C2 if Za <Z _R × 0.5)
Becomes If it is), it is determined to be a non-defective product, the F1 counter is incremented, and the secondary determination is completed.

If Z _max -Z _min > C, then a 7 × 7 orbital pixel group (filter F3) is selected as a defect candidate.
Are set as indicated by the diagonal lines in FIG. After this setting, non-defective products are determined by the same processing as when the filter F1 is set. Here again, if there is no determination of non-defective products, a 5 × 5 orbital pixel group (filter F
2) is set as indicated by the diagonal lines in FIG. 4B, and the non-defective product is determined in the same manner as described above. If it is not determined to be the non-defective product, it is determined to be defective in the secondary determination and the NG counter is counted. Up and output the judgment result.

Here, the filters F2 and F3
The reason why the corner pixel is not included in the orbiting pixel group is because the distances to the respective circuit edges of the XY direction pattern and the diagonal direction pattern are made substantially equal to each other. Further, in the above determination process, the uniformity reference of each orbiting pixel group is set to the same value, but different values may be used. The order of determination may be different depending on the selection and combination of the circulating pixel groups. Here, usually, the frequency of over detection is small, but the frequency is high, and it is more efficient and faster to determine the size of the circular pixel group first. In addition, since the second method removes low-density pseudo defects, it is efficient to make a determination using a large surrounding pixel group.

FIG. 6 shows the secondary judgment result of the portion corresponding to the circuit pattern of FIG. 2. In the figure, the pixels painted in black are the defect judgment pixels in the secondary judgment, and large defect candidate pixels are reduced. Yes, small defects that are far from the circuit can be erased. As described above, since it is possible to reduce the excessive number of defect candidate pixels, it is possible to efficiently perform final pass / fail judgment and defect information output (indicating a position and a type) as an inspection device. Also, the number of pixels deleted by the filters F1, F2, F3 of the orbiting pixel group and the number of pixels as the final defect are displayed on the monitor screen (not shown) according to the counter values of F1, F2, F3, and NG, and the determination status is displayed. Is easy to understand. That is, the operator can understand not only the final NG value but also the occurrence status of pseudo defects, and the large number of pseudo defects is a problem in process control, and can be utilized as information for improving process quality.

The F1, F2, F3, and NG counters are counters built in the secondary determination unit 13. In addition to the method of the secondary determination described above, the reference value of the selection, combination, order or uniformity of the applicable surrounding pixel groups may be changed according to the area division data such as the area A shown in FIG.

In this example, the positional accuracy such as the mounting hole H,
The area A is set to a portion where the shape is unstable and is easily detected excessively, and a portion where some defects such as a power source and a ground solid portion (not shown) are not a problem, and the inspection standard is loosened significantly. However, not a non-inspection, but a required level of defect inspection is performed, and it is a great feature that the defect detection is surely detected particularly for abnormal or unexpected abnormality.

Usually, in such a case, in order to avoid over-detection, in many cases, a mask is simply used to make such an area uninspected, which makes it vulnerable to unexpected defects. In the above-mentioned secondary determination, the determination is made based on the uniformity of the revolving pixels. However, as another example, any pixel of the detection pattern data of the revolving pixels is within the data range of the circuit portion, and any other pixel is not It is also possible to check whether the data is within the data range of the circuit portion, judge as defective only when both are satisfied, and make the other non-defective.

As another example, the reference data of the revolving pixel is read out, and for example, it is the circuit part, the non-circuit part, or the mounting hole H from the upper limit reference value L ₁ or the lower limit reference value L _2.
It may be possible to make a secondary determination by discriminating between the specific portion such as the above and the boundary portion thereof. Moreover, you may combine these as another example. As another example of the device, FIG.
The inspection method of the present invention may be used as a method of detecting a circuit pattern by chromaticity by using a color camera as the image pickup camera 1. In this case, the pseudo defect of the circuit pattern, dirt,
It is possible to reduce excessive detection due to the color of foreign matter.

Further, as shown in FIG. 7, in the height detecting device 15 such as a light cutting method for the image pickup means I, or in the height detecting method by automatic focusing, if the inspection method of the present invention is used, dust, flaws, and circuit patterns can be obtained. It is possible to reduce the excessive detection due to the shape variation and the like. Further, as shown in FIG. 8, the X-ray source 17 irradiates the inspected circuit board 4 with X-rays, and the X-ray transmission amount transmitted through the inspected circuit board 4 is determined by the X-ray imaging apparatus 1.
The method of the present invention may be adopted in an apparatus for inspecting a molded circuit pattern or a circuit pattern inside a multilayer laminate using the image pickup means I detected in 6. In this case, in addition to the defect of the circuit pattern, a defect such as a metal foreign matter defect in the inside becomes a problem. However, it is an excessive detection to detect a defect that is distant from the circuit pattern portion as a defect. Can also be reduced.

[0045]

According to the first aspect of the present invention, there is provided a circuit pattern inspection method for detecting a defect in a circuit pattern by comparing detected pattern data created by imaging a pattern to be inspected with reference data. The primary determination process of detecting the defect correction by comparing with the reference data in which the allowable range of the detection pattern data is set, and the pixel determined to be the defect correction in the primary determination process, the predetermined arrangement centered on the pixel The invention according to claim 10 further comprises a secondary determination process of setting a revolving pixel group and determining defect / non-defect based on the detection pattern data of the revolving pixel group. Means, a detection pattern data section for storing the detection pattern data output from the image pickup means, and a reference data set with an allowable range of the detection pattern data. A reference data creating unit, a reference data unit that stores the reference data created by the reference data creating unit, a primary determining unit that detects the defect candidate by comparing the detection pattern data with the reference data, and For a pixel determined as a defect candidate by the determining means, a revolving pixel group having a predetermined arrangement centered on the pixel is set, and a secondary determining means for determining a defect / non-defect based on detection pattern data of the revolving pixel group is provided. Therefore, it is possible to determine the defect correction of the primary judgment by matching the distance to the circuit edge part on the circuit pattern and the size of the defect candidate, and it is unnecessary to exclude unnecessary defect correction (pseudo defect). The number of detections can be reduced, and for a large defect, the number of internal defective defect correction pixels can be reduced and only a part of pixels can be narrowed down, whereby data storage and post-processing can be efficiently performed. In this way, the reliability of the inspection is increased, the quality is stabilized, the work of reviewing defective products is reduced, labor is saved, and productivity is improved. Further, there is an effect that the inspection process can be inlined, which leads to a great improvement in productivity of the entire circuit board manufacturing line. In particular, the invention of claim 1 relates to
Multiple types of fixed placement are provided for each pixel group size,
Until the judgment is made, from the first orbital pixel group to the last orbital image
Since the secondary judgment is repeated sequentially up to the prime group, the size of defect compensation is large.
The allowable distance from the edge of the circuit pattern
By making a distinction, even if a relatively large pseudo defect
It can be non-defective if it is far from the edge of the pattern.
Which in turn reduces over-detection and improves productivity
There is an effect. Further, the invention of claim 10 is the secondary judgment hand.
The step has a predetermined fluctuation range of the detection pattern data of the revolving pixel group.
Since it is a means to determine non-defect within the uniformity of
It is easy and suitable for high-speed processing, and can perform efficient and high-speed inspection.
There is an effect.

[0046]

ClaimsTwoOf the invention is the smallest circumference
Since the rotation pixel group is the first revolving pixel group, a small pseudo
Even if there are many defects, high-speed processing can be performed efficiently, and
In this case as well, the number of defective defective pixels inside is reduced, so this case is also effective.
The effect is that high-speed processing can be performed efficiently. ClaimThree
Is the detection pattern of the orbiting pixel group in the secondary determination process.
If the fluctuation range of data is within the prescribed uniformity, it is judged as non-defect
Easy, suitable for high-speed processing, efficient and high-speed inspection
There is an effect that can be.

The invention of claim 4 is to more sets of predetermined uniformity, the large <br/> can the average value of the detected pattern data of the orbiting pixel group is used to switch the predetermined uniformity ,
The parts on the circuit pattern can be identified, and the uniformity that suits each one
There is an effect that it is possible to accurately determine non-defects based on the standard , and further reduce excessive detection.

According to the fifth aspect of the invention, in the secondary determination process, if the detection pattern data of the revolving pixel group includes both the data of the circuit portion and the non-circuit portion of the circuit pattern to be inspected, it is determined to be defective. With a simple algorithm, it is possible to recognize a defect candidate near the circuit edge of the circuit pattern and to perform highly reliable inspection.

According to a sixth aspect of the present invention, it is determined from the reference data of the circulating pixel group whether the circuit portion, the non-circuit portion, or the specific portion of the circuit pattern to be inspected or the boundary portion thereof is set, and a plurality of types are set. According to the invention of claim 7 , the circuit portion, the non-circuit portion, or the specific portion of the circuit pattern to be inspected is selected from the reference data of the circulating pixel group. to determine if it is a boundary portion of the orbiting pixel groups, since selecting and using a predetermined uniformity of the detection pattern data fluctuation range set more, further the invention of claim 11, the secondary determination unit It is determined whether the circuit portion, the non-circuit portion, or the specific portion of the circuit pattern to be inspected or the boundary portion thereof is determined from the reference data of the orbiting pixel group, and a predetermined arrangement of a plurality of orbiting pixel groups is set. Is a means for selecting and using a predetermined uniformity of the fluctuation width of a plurality of set detection pattern data, it is possible to accurately determine the defect occurrence site on the circuit pattern, and to make a determination according to each site. This has the effect of further reducing excess detection.

According to the eighth aspect of the present invention, the area division data for dividing the inspection area which is created in advance is used, and the predetermined area of the plurality of kinds of the surrounding pixel groups is set according to the area division data of the pixel judged as the defect candidate in the primary judgment process. Since the arrangement is selected and used,
The invention according to claim 9 uses the area division data that divides the inspection area that is created in advance, and uses the area division data of the pixels that are determined to be defect candidates in the primary determination process, and sets a plurality of types of detection pattern data for the revolving pixel group. Since the predetermined homogeneity is selected and used, the invention according to claim 12 further comprises an area division data section which creates area division data for dividing the inspection area, and stores the area division data, and the secondary determination means is the primary determination means. The area division data of the pixel determined by the determination means is a means for selecting and using a predetermined arrangement of the orbiting pixel group having a plurality of types set or the predetermined uniformity set for a plurality of types. It is possible to set, and therefore it is possible to judge pass / fail by a separately determined criterion for a specific area of the circuit pattern, it is possible to deal with exceptional situations, and it is possible to reduce these excessive detections. There is.

[0052]

[0053]

Therefore, as described above, since the circuit pattern inspection method and the inspection apparatus thereof according to the present invention can reduce the excessive detection, the reliability of the inspection is improved, the quality is stable, and the review work can be reduced. Labor saving will be promoted and productivity will be improved. Further, the inspection process can be made in-line, and the productivity of the entire circuit board manufacturing line can be greatly improved.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an entire apparatus according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of the primary determination of the above.

FIG. 3 is an explanatory diagram of a circular pixel group of the above.

FIG. 4 is an explanatory diagram of filter setting for primary determination in the above.

FIG. 5 is a flowchart of secondary determination according to the above.

FIG. 6 is an explanatory diagram of secondary determination according to the above.

FIG. 7 is a configuration explanatory view of an image pickup means of another embodiment of the above.

FIG. 8 is a configuration explanatory diagram of an image pickup means of another embodiment of the above.

[Explanation of symbols]

1 Imaging camera 2 NC drive 3 tables 4 Inspected circuit board 5 Lighting lamp 6 A / D converter 7 Lighting section 8 Imaging control unit 9 Detection pattern data section 10 Standard data creation section 11 Standard data section 12 Primary judgment section 13 Secondary judgment section 14 Area division data section

Claims (12)

(57) [Claims] 1. A method of inspecting a circuit pattern for detecting a defect in the circuit pattern by comparing detection pattern data created by imaging a circuit pattern to be inspected with reference data, and allowing the detection pattern data and the detection pattern data. For the primary judgment process of detecting the defect correction by comparing with the reference data in which the range is set, and for the pixel judged to be the defect correction in the primary judgment process, set the surrounding pixel group of a predetermined arrangement centered on the pixel However , it has a secondary judgment process for judging a defect / non-defect based on the detection pattern data of the circulating pixel group.
Multiple types of pixel groups can be arranged according to the size of the pixel group.
Until the non-defect is determined, the first
A method of inspecting a circuit pattern, characterized in that secondary determination is sequentially repeated until a final revolving pixel group . 2. A circular pixel group having the smallest size
2. The circuit pattern inspection method according to claim 1, wherein the circuit pixel group is a group of circular pixels . 3. A detection of a group of revolving pixels in the secondary determination process.
If the fluctuation range of the pattern data is within the prescribed uniformity, it is
The method for inspecting a circuit pattern according to claim 1 or 2, wherein it is determined that a defect occurs . 4. A group of revolving pixels in which plural kinds of predetermined uniformity are set.
According to the size of the average value of the detection pattern data of
4. The uniformity of the above is switched and used.
Inspection method of the described circuit pattern. 5. A circuit portion of a circuit pattern to be inspected, wherein detection pattern data of a group of revolving pixels is used in the secondary determination process.
And the inspection method of a circuit pattern according to claim 1, wherein that you determined defective if, including both data of the non-circuit portion. 6. An object to be inspected based on the reference data of the orbiting pixel group.
Circuit part, non-circuit part, or specific part of the circuit pattern,
Alternatively, it is determined whether these are the boundary parts and multiple types are set.
Method of inspecting a circuit pattern according to any one of claims 1 to 5, wherein the selectively using predetermined arrangement of boss was orbiting pixel group. 7. An object to be inspected based on the reference data of the orbiting pixel group.
Circuit part, non-circuit part, or specific part of the circuit pattern,
Alternatively, it is determined whether or not these are the boundary parts, and
Specified fluctuation range of detection pattern data for multiple groups
Claim 1乃, wherein the selectively using of uniformity
7. A method for inspecting a circuit pattern according to any one of items 6 to 6 . 8. Area division for dividing an inspection area created in advance
Pixels judged to be defect candidates in the primary judgment process using data
The area classification data of the
The predetermined arrangement is selected and used.
8. A circuit pattern inspection method according to any one of 7 to 7 . 9. Area division for dividing an inspection area created in advance
Pixels judged to be defect candidates in the primary judgment process using data
By setting the area classification data of
The method for inspecting a circuit pattern according to any one of claims 1 to 7, wherein a predetermined uniformity of the detected pattern data is selected and used . 10. Detecting detection pattern data of an inspection target
Image capturing means and a detection pattern output from the image capturing means.
Detection pattern data section that stores the pattern data and the detection pattern
Create reference data that sets the allowable range of turn data.
Created by the reference data creation unit and the reference data creation unit
A reference data section for storing reference data, and the detection pattern
Defect data by comparing the reference data with the reference data
The primary judgment means and the primary judgment means judged the defect candidate.
A pixel that is a revolving pixel with a predetermined arrangement centered on the pixel
Group is set, and by the detection pattern data of the orbiting pixel group
Equipped with secondary judgment means for judging defects / non-defects
The means determines the fluctuation range of the detection pattern data of the orbiting pixel group.
An apparatus for inspecting a circuit pattern, which is means for determining non-defects within a predetermined uniformity . 11. The secondary determination means is a reference data for the orbiting pixel group.
Circuit part and non-circuit part of the circuit pattern to be inspected
Minutes, or a specific part or a boundary part of them.
Separately, a predetermined arrangement or multiple
Specified uniformity of fluctuation range of set detection pattern data
Characterized in that it is a means to select and use the claim 1
A circuit pattern inspection device described in 0. 12. Area division data for dividing an inspection area is created.
It is equipped with an area classification data section that stores this
The determining unit determines the area of the pixel determined to be the defect candidate by the primary determining unit.
Predetermined number of orbital pixel groups set by multiple types based on area classification data
Fluctuation range of detection pattern data that has been arranged or multiple
12. The circuit pattern inspection apparatus according to claim 10, which is a means for selecting and using a predetermined uniformity of
Place