JP2001056301A - Circuit pattern inspecting method and device to be used therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine the suitability of a circuit pattern by obtaining the amounts of displacement in X- and Y-directions for every location correction area of master image data, shifting the master image data in the X- and Y-directions by the obtained amounts, and comparing it with image data for inspection for every location area. SOLUTION: A line memory+shift register 52 for developing master data develops image data captured by dividing master data for every location correction area into M×N pieces. The developed data is transmitted to a mismatch enumerating circuit 54 to obtain the number of mismatches with defect candidate data, and the obtained number is stored in enumeration result storage memory 56. The location correction value of each location correction area is obtained at an area-by-area location correction value calculating routine 58 and compared with the location correction value of an adjacent location correction area at a location correction modification routine 60, and a location correction value in the closest area is substituted at the time of separation by a predetermined value or more. The master data is shifted in the X- and Y-directions by the location correction value determined for every location correction area and compared with image data for inspection for every location correction area to determine the GO/NO-GO of a circuit pattern.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a circuit pattern inspection method and an apparatus for inspecting whether or not a pattern is properly formed in a process of manufacturing a printed wiring board.

[0002]

2. Description of the Related Art In order to cope with a higher density of printed wiring boards, inspections are currently being performed by optically processing a wiring pattern photographed by a camera. As this inspection method, a feature extraction method and a complete comparison method are mainly used. Here, the feature extraction method uses features extracted from non-defective boards, and the complete comparison method uses image data of non-defective boards.
It is used as master data for each test.

[0003]

However, when an inspection is performed by using one master data with a plurality of inspection machines, there is a mechanical error (mechanical error) between the inspection machines, There is also an error (optical system error) in the settings (focus etc.) between the provided cameras. Here, with the increase in density of printed wiring boards, it is very difficult to adjust these mechanical errors and optical system errors, that is, to adjust a plurality of inspection machines so that the same conditions are satisfied. For this reason, by enlarging the above master data, mechanical errors and optical errors are accommodated, but enlarging the master data effectively expands the width considered as non-defective products and overlooks serious defects. Cause.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problem, and an object of the present invention is to use a single master data to properly print a printed wiring board with a plurality of circuit pattern inspection apparatuses. It is an object of the present invention to provide a circuit pattern inspection method that can be inspected at a high speed, and a circuit pattern inspection apparatus used for the method.

[0005]

According to a first aspect of the present invention, in order to achieve the above object, a plurality of circuit pattern inspection apparatuses use one piece of master image data to compare the acquired image data with inspection image data. A circuit pattern inspection method for judging the suitability of a circuit pattern by the following (A)
A circuit pattern inspection method including a deviation amount measuring step and (B) a propriety judging step: (A) a step of measuring a deviation amount between master image data and an inspection device, which has the following (1) to (3); 1) The master image data is represented by m (Y direction) × n
(2) The image data of the acquired position correction area is expanded into M × N image data, and (3) The developed data and the data obtained by dividing the inspection image data into the position correction areas are referred to as M ×
Y for each position correction area compared with each of N image processing units
The shift amount in the direction and the X direction is determined. (B) The master image data is shifted in the Y direction and the X direction for each position correction area by the measured deviation amount, and is compared for each position correction area of the acquired inspection image data to determine whether the circuit pattern is appropriate. Process.

According to a second aspect of the present invention, in the first aspect, the position correction area is divided into a plurality of portions corresponding to one interruption interval of a control device of image data taken in from a camera.

According to a third aspect of the present invention, in the first aspect, the deviation amount is determined by counting the number of coincidences or non-coincidences of M × N image processing units, and determining the Y position at the position having the largest number of coincidences.
It is a technical feature that the shift amount in the direction and the X direction is determined as a shift amount for each position correction area.

According to a fourth aspect, in the first to third aspects, when the position having the largest number of matches cannot be specified or the number of matches is smaller than a predetermined value in the determination of the shift amount in the above (3). The technical feature is to use, as the shift amount of the position correction area, the shift amount of an adjacent position correction area for which the shift amount can be determined.

According to a fifth aspect of the present invention, in the first to fourth aspects, in the determination of the shift amount in (3), after a position having a large number of coincidences is determined, the determined shift amount is shifted by an adjacent position correction area. It is a technical feature that it is determined whether the position correction area is larger than a predetermined value by comparing with the amount, and when the position correction area is larger, a deviation amount of an adjacent position correction area for which the deviation amount can be determined is used as the deviation amount of the position correction area. .

According to a sixth aspect of the present invention, in the second aspect, when it is determined in (B) that the circuit pattern is inappropriate, the following steps (5) to (8) are performed. Features: (5) The master image data is represented by m2 (Y direction) × n
A second position correction area for performing position correction consisting of 2 (X direction) image processing units, and for each second position correction area consisting of one interrupt interval of the control device for image data taken in from the camera. Dividing and importing,
(6) a step of developing the captured image data of the second position correction area into M × N images; (7) dividing the expanded data and the inspection image data into a second position correction area and capturing the divided data; Determining the amount of deviation in the Y direction and the X direction for each second position correction area by comparing the data with each other in the M × N image processing units; The image data is converted into Y for each second position correction area.
Determining the suitability of the circuit pattern by comparing the acquired image data for inspection in each of the second position correction areas by shifting in the direction and the X direction.

A seventh aspect of the present invention is a circuit pattern inspection apparatus which determines whether a circuit pattern is appropriate by comparing master image data with captured inspection image data. (Y direction) × n
(X direction) The image data of the position correction area is divided into each position correction area for performing position correction consisting of image processing units in the (X direction), and the obtained image data of the position correction area is developed into M × N images. For each of the M × N image processing units, data obtained by dividing the image data for position correction into the position correction areas is compared for each of the M × N image processing units, and a value that determines the amount of deviation in the Y direction and the X direction for each position correction area is held. (B) a shift amount holding unit;
The master image data is shifted in the Y direction and the X direction by the measured deviation amount for each position correction area, and is compared for each position correction area of the acquired inspection image data to judge the suitability of the circuit pattern. A technical feature.

According to the first aspect of the present invention, the amount of deviation of the master image data in the Y direction and the X direction is determined for each position correction area. Then, the master image data is shifted in the Y direction and the X direction by the determined shift amount for each position correction area,
A comparison is made for each position correction area of the captured inspection image data to determine whether the circuit pattern is appropriate. For this reason, even if there is a mechanical error and an optical system error between the master data and the circuit pattern inspection device that has created the master data and the circuit pattern inspection device that performs the inspection, the master image data and Is shifted in the Y direction and the X direction for each position correction area, and the printed circuit board can be accurately inspected by a plurality of circuit pattern inspection apparatuses using the master data.

According to the second aspect of the present invention, the position correction area is
Master data is corrected not only in the XY direction but also in the rotation direction for the image data captured from the camera, because one interrupt interval of the control device for the image data captured from the camera is divided into a plurality of units. can do.

According to the third aspect of the present invention, the amount of deviation is determined by M
The number of matches or mismatches of the × N image processing units is counted, and the shift amount in the Y direction and the X direction to the position having the largest number of matches is determined as the shift amount for each position correction area. Can be determined.

According to the fourth aspect of the present invention, when the position having the largest number of matches cannot be specified in determining the shift amount, the shift amount of the position correction area is determined as the shift amount of the adjacent position correction area for which the shift amount can be determined. Since the amount is used, an appropriate shift amount can be determined for all the position correction areas.

According to a fifth aspect of the present invention, in the determination of the shift amount, after the position having the large number of coincidences is determined, the determined shift amount is larger than the shift amount of the adjacent position correction area by a predetermined value or more. When the determined shift amount is inappropriate, the shift amount of the adjacent position correction area for which the shift amount can be determined is used as the shift amount of the position correction area. The amount of misalignment can be determined.

According to a sixth aspect of the present invention, a circuit pattern is inspected by using master data in which the amount of displacement is corrected in units of position correction areas, that is, master data in which mechanical errors and optical errors are measured and position correction is performed. When it is determined that the printed circuit board is inappropriate, the position of the master data is corrected in real time, and the circuit pattern is inspected again. It can be inspected properly.

In the present invention, the amount of deviation of the master image data in the Y direction and the X direction is determined for each position correction area.
Then, the master image data is shifted in the Y direction and the X direction for each position correction area by the determined shift amount, and is compared for each position correction area of the captured inspection image data.
Determine whether the circuit pattern is appropriate. For this reason, even if there is a mechanical error and an optical system error between the master data and the circuit pattern inspection device that has created the master data and the circuit pattern inspection device that performs the inspection, the master image data and Is shifted in the Y and X directions for each position correction area to correct the position.
The printed wiring board can be inspected accurately.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A circuit pattern inspection apparatus and a circuit pattern inspection method according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of the circuit pattern inspection apparatus according to the first embodiment. The circuit pattern inspection device 10 is a printed wiring board (not shown) to be inspected.
, A servomotor 14 for adjusting the position of the inspection table 12, an illumination 16 for irradiating the printed wiring board with light, a CCD camera 18 for photographing the printed wiring board, and a CCD camera 18. An A / D converter 20 for converting photographed image data into digital values,
A binarizing unit 22 for binarizing the image data converted into digital values, and a feature extracting unit 2 for extracting features in the image data
6 and a switch 28 for distributing image data
A master data memory unit 36 for holding master data; a defect matching unit 30 for comparing the master data held in the master data memory unit with the captured image data to inspect for defects; The apparatus includes a position correction unit 29 for performing correction, a data totalizing unit 32 for totalizing the results of inspection performed by the defect matching unit 30, and a terminal (monitor) 34 for displaying inspection results. Switch 28
Is connected to the master data memory unit 36 in the “master data creation mode” for creating master data, and is connected to the position correction unit 29 in the “inspection mode” for inspecting the circuit pattern. Connected. The circuit pattern inspection device 10 is connected to a defect recognition device 40 for allowing an operator to recognize a defective portion.

FIG. 2 is a block diagram showing processing in the defect collating unit 30, the master data memory unit 36, and the data editing unit 32. In the figure, the left side is a block configured by hardware 50A, and the right side is a block configured by software 50B. In the circuit pattern inspection device 10 of the first embodiment, the master data is shared by a plurality of circuit pattern inspection devices 10. For this reason, a position error between the master data and each circuit pattern inspection apparatus 10 is measured and stored in units of a position correction area described later. Then, based on the measured position error,
The inspection is performed by correcting the position of various master data corresponding to the inspection target in units of the position correction area. Here, the upper half in FIG. 2 is a position correction mode for measuring and holding the position error in units of a position correction area, which is performed before the start of the inspection, and the measurement result is stored in the hard disk drive 62.
Is written to. On the other hand, the lower half is an inspection mode for inspecting a circuit pattern based on the measured position error. In addition,
The above-described master data for measurement does not need to be special, and the error measurement can be performed using the master data for inspection.

First, the position correction mode will be described.
Here, in the present embodiment, the position is corrected for each position correction area. The configuration of this position correction area is shown in FIGS.
Shown in FIG. 3 shows a configuration of image data (inspection area) captured from the camera 18. This path corresponds to the scanning width of the camera 18, and one path corresponds to three position correction areas. On the other hand, the inspection direction corresponds to the scanning direction of the camera, and one interruption interval of the CPU (not shown) of the circuit pattern inspection device corresponds to the inspection direction component of one position correction area. That is, at one interrupt interval,
The image data for the three position correction areas is captured.

In this embodiment, the master data is corrected by measuring the correction amounts (shift amounts) in the X and Y directions for each position correction area, as will be described later. Since one interrupt interval of image data taken from the camera is divided into three, the master data can be corrected not only in the XY direction but also in the rotation direction in units of position correction areas in the rotation direction. it can.

FIG. 4 shows the configuration of the position correction area shown in FIG. The number of the position correction areas is m in the Y direction,
It consists of n master data cells in the X direction. Each master data cell includes a minimum image processing unit for extracting a feature of a circuit pattern during image processing.

The description of the position correction mode will be continued with reference to FIG. First, the master data is obtained by the master data XY development line memory + shift register 52.
The image data of the position correction area divided and taken in for each position correction area is developed into M × N image data. The master data XY development line memory + shift register 52 includes an M-stage line memory 52A connected in series as shown in FIG. 5, and an N-stage shift register 52B connected to each line memory 52A. .

Next, the processing in the mismatch counting circuit 54 will be described. The non-coincidence counting circuit 54 includes the master data XY expansion line memory
× N image data expanded and defect candidate data (data captured from the camera 18 and subjected to feature extraction)
Is counted. This mismatch counting circuit 54
Comprises M × N mismatch counting blocks shown in FIG. As shown in FIG. 6 (A), the master data provided to the mismatch counting circuit 54 is obtained from the master data expanded in M cells in the master data unit in the Y direction and N cells in the X direction in the master data unit. M × N pieces are cut out from a window having the same size as the master data. That is, M × N master data are sent to the respective mismatch counting blocks.

The mismatch counting block comprises a logic circuit 54A and an S-bit counter 54B, counts the places where the output of the logic circuit 54A becomes high, and outputs the result as S-bit data. The location where the logic circuit 54A goes high is
This is where there is defect candidate data even though there is no master data. In other words, the mismatch counting block detects a defect in the corresponding master data (master data shifted in the X direction and the Y direction by the XY development line memory + shift register 52) for each master data cell. The count of mismatch with the candidate data is output.

Subsequently, the counting result storage memory 56 holds the number of mismatches counted in the M × N mismatch counting blocks.

Next, in an area-specific position correction value calculation routine 58, a position correction value is calculated in master data units in all position correction areas included in the inspection area. In the area-based position correction value calculation routine 58,
The minimum mismatch number is obtained from the M × N mismatch numbers held in the counting result storage memory 56, and a position (shift amount in master data cell units) where the mismatch number is minimized is searched for. Is specified as a position correction value, and if it cannot be specified, the position correction value is not obtained and thus invalidated.

The processing in the area-specific position correction value calculating routine 58 will be described in more detail with reference to the flowchart of FIG. First, a loop is set for all areas (S12). Next, all the mismatch numbers obtained by the mismatch counting circuit 54 and held in the counting result storage memory 56 are obtained (S14). Then, a search is made for the minimum value among the acquired mismatch numbers (S16). Subsequently, a cell in which the number of mismatches is equal to the minimum value is searched (S1).
8). Finally, if you can identify the searched cell, for example,
It is determined whether there is one cell having the minimum value and whether the minimum value is equal to or greater than a predetermined threshold value (S20). If there are a plurality of cells having the minimum value, or the minimum value of the number of mismatches is the threshold value If it is larger, the process proceeds to step 24, where the information of the searched cell is discarded, and the position correction value is invalidated. On the other hand, if the number of cells having the minimum value is one and the minimum value of the number of mismatches is smaller than the threshold value, the process proceeds to step 22 and information of the searched cell is registered as a position correction value.
For example, as a position correction area of the master data cell,
When the number of inconsistencies of the cells shifted by 3 in the X direction and 4 in the Y direction by the master data cell unit is “2”, which is the minimum, the position correction value (shift amount) of the position correction area will be described later. Thus, 3 is set in the X direction and 4 is set in the Y direction. When the above processing is performed for all the position correction areas, the routine processing ends.

Subsequent to the area-specific position correction value determining routine 58, a position correction correction routine 60 is executed. In the position correction correction routine 60, first, the value of the position correction area for which a valid position correction value has been obtained is compared with the value of the position correction area of the closest position to determine whether the position is separated by a certain cell or more. Judge whether or not. This is performed to ensure continuity of the position correction value. Next, the value of the closest area is substituted for the position correction area where the position correction value is determined to be invalid.

The processing in the position correction / correction routine 60 will be described in more detail with reference to the flowchart of FIG. First, a loop is set for all areas (S
30). Next, the position correction value of the position correction area for determining the position correction value is loaded (S32), and it is determined whether the position correction value is valid or invalid (S34). If the position correction value is invalid, the process returns to step 32 and loads the position correction value of the next position correction area to be processed. On the other hand, if the position correction value is valid, the process proceeds to step 36, where the position correction value of the position correction area closest to the position correction area to be corrected is loaded, and the position correction value of the position correction area to be determined is loaded. A position correction value of the adjacent position correction area is compared (S38). Here, when the position correction value of the position correction area to be determined and the loaded position correction value are not separated by a predetermined value P or more, it is determined that the position correction value of the position correction area is appropriate, and the process returns to step 32. . On the other hand, when the position is separated by the predetermined value P or more, for example, when processing the position correction area indicated by oblique lines in FIG. 3, the position correction of the closest position correction area (here, No. 1 in the figure) is performed. Values 3 and 4 are the position correction values 5 and 5 of the position correction area of the diagonal line, when the distance is more than P (2 in the Y direction and 1: here, P is 2 in the X direction) and there is no continuity Determines that the position correction value of the position correction area is an inappropriate value, shifts to step 42, and invalidates the registered position correction value.

When the processing of steps 32 to 42 is looped for all areas, a loop for all areas is set again at step 40. Then, the position correction value of the position correction area for correcting the position correction value is loaded (S4).
4), it is determined whether the position correction value is valid or invalid (S4)
6). Here, if the position correction value is valid, the process returns to step 40, and the process proceeds to the next position correction area. On the other hand, if the position correction value is invalid,
Then, the process proceeds to step S8, and the position correction value of the position correction area closest to the position correction area to be corrected is loaded. Here, “closest” means that when processing the position correction area indicated by oblique lines in FIG. 3, the first position correction area is the closest, and the position correction value (3 , 4) is valid, the position correction value (3, 4) is loaded. On the other hand, when the first position correction value is not valid, the position correction value of the second position correction area is loaded, and when the second position correction value is not valid, the position correction value of the third position correction area is loaded. . 1 in the figure
Reference numeral 17 denotes a position correction area numbered in the order of proximity. Then, the loaded closest position correction value is copied to the position correction area to be corrected (S50).
Here, the first position correction value (3, 4) is used for the hatched position correction area. The above process is looped for all areas.

In the above-described processing, a position correction value is given to all the position correction areas, and the given position correction value is written into the hard disk drive 62, and the processing in the position correction mode ends.

In the present embodiment, when the position cannot be specified in the area-based position correction value determining routine 58 described above with reference to FIG. 8, the proximity correction for which the shift amount can be determined is used as the position correction value of the position correction area. Since the position correction values of the position correction areas to be used are used, it is possible to determine appropriate position correction values for all the position correction areas.

Further, in the present embodiment, after the position correction value is determined in the above-described area-based position correction value determining routine 58, the determined position correction value is compared with the position correction value of the adjacent position correction area to determine a predetermined value. When the position correction value is larger than the value, that is, when the determined shift amount lacks continuity and is inappropriate, the position correction value of the adjacent position correction area for which the position correction value can be determined is determined as the position correction value of the position correction area. Is used, an appropriate position correction value can be determined for all position correction areas.

Next, an inspection mode for actually inspecting a circuit pattern will be described. Using the position correction value determined in the position correction mode, the master data (there is no need to be the master data measured above) is position corrected for each position correction area, and the captured image data (defect candidate data) ). Here, first, a position correction value setting routine 64 for reading a position correction value from the hard disk drive 62 and setting the same in the position correction value register 66 will be described.

FIG. 1 shows the position correction value setting routine.
This will be described in more detail with reference to the flowchart of FIG. First, loops for all areas are set (S60). Next, a position correction value (coordinate data) for each position correction area is read from the hard disk drive 62 (S62), and the read coordinate data is set in the position correction value register 66 (S64).

Next, the processing in the inspection mode on the hardware 50A side will be described. Here, first, the configuration of the matching circuit 68 with the master data correction function will be described. As shown in FIG. 7, the matching circuit 68 with the master data correction function includes an M-stage line memory 68A, a Y-axis selector 68Y, an N-stage shift register 68B, an X-axis selector 68X, and a logic circuit 68C. It consists of. First, the master data is stored in an M-stage line memory 68A.
Are expanded to M levels, and Y-axis selector 68Y
The master data corrected by the axis position correction value is selected. For example, when “3” is set as the position correction value on the Y-axis side in the position correction area of the diagonal line described above with reference to FIG. 3, “3” is set in the Y-axis direction in master data cell units.
The shifted master data is selected, and the master data shifted by “3” is stored in an N-stage shift register 68B.
, The data is developed in N stages in the X-axis direction, and the master data corrected by the X-axis position correction value by the X-axis selector 68X is selected. For example, when “4” is set as the position correction value on the X axis side, master data shifted by “4” in master data cell units in the X axis direction is selected, and “3”, X in the Y direction are selected. Master data shifted in the direction "4", that is, Y for each position correction area based on the position correction value.
And the master data whose position has been corrected in the X-axis direction is output. Then, the master data output from the X-axis selector 68X and the defect candidate data fetched from the camera side are compared by the logic circuit 68C. If the master data is not the master data but exists on the defect candidate data side, Outputs high. Then, this high is written to the defect candidate storage memory 72. The defect coordinate counter 70 is an auxiliary counter for obtaining coordinates.

The data written in the defect candidate storage memory 72 is processed into data for displaying a defect on the monitor 34 in a defect information processing routine, and then output.

In the circuit pattern inspection apparatus and the inspection method according to the first embodiment, the shift amounts of the master data in the Y direction and the X direction are determined for each position correction area. Then, the master image data is shifted in the Y direction and the X direction for each position correction area by the determined amount of deviation, and is compared for each position correction area of the acquired inspection image data to determine whether the circuit pattern is appropriate. For this reason, even if there is a mechanical error and an optical system error between the master data and the circuit pattern inspection device that has created the master data and the circuit pattern inspection device that performs the inspection, the master image data and Is shifted in the Y direction and the X direction for each position correction area, and the printed circuit board can be accurately inspected by a plurality of circuit pattern inspection apparatuses using the master data.

Next, a circuit pattern inspection apparatus and an inspection method according to a second embodiment of the present invention will be described. In the second embodiment, when the circuit pattern is inspected by the method of the above-described first embodiment, when a defect of the circuit pattern is detected, the method is simplified again for further confirmation. Inspection is performed by processing. Note that the configuration of the circuit pattern inspection device of the second embodiment is the same as that of the first embodiment described above with reference to FIG. 1, and therefore the description thereof will be omitted with reference to FIG. 1.

That is, in the first embodiment, an error from the master data is obtained in advance, and the obtained position complement value is stored in the hard disk drive 62. On the other hand, in the simplification processing of the second embodiment, the position correction is performed by comparing the master data and the defect candidate data in real time, and the circuit pattern is inspected.

The capture of the position correction area in this simple processing will be described with reference to FIG. As described above with reference to FIG. 3, in the circuit pattern inspection apparatus according to the first embodiment, the image data for one interrupt is divided into three in the inspection (scanning) direction of the camera to obtain a position correction area. In this simple processing, image data for one interrupt is used without division.

FIG. 11 is a block diagram showing processing in the defect collating unit 30, the master data memory unit 36, and the data editing unit 32 in the simplified processing of the second embodiment. Hard disk drive 6 for real-time correction
The second embodiment is the same as the first embodiment described above with reference to FIG. 2 except that the step 2 is omitted and the contents of the position correction value correction routine 160 are simplified. Therefore, only the contents of the position correction value correction routine 160 are shown in FIG.
This will be described with reference to the flowchart of FIG.

First, a position correction value of a position correction area for determining a position correction value is fetched (S70). Then, it is determined whether the position correction value is valid or invalid (S72). Here, if the position correction value is invalid, the process proceeds to step 80, where 1
The position correction value of the immediately preceding position correction area (see FIG. 10) is substituted. On the other hand, if the position correction value is valid, the process proceeds to step 74, where the position correction value of the position correction area immediately before the position correction area to be corrected is fetched, and the position correction value of the position correction area to be determined is fetched. A position correction value of the immediately preceding position correction area is compared (S76). Here, when the position correction value of the position correction area to be determined and the captured position correction value are not separated by a predetermined value P or more, it is determined that the position correction value of the position correction area is appropriate and the process returns to step 71. . On the other hand, when it is separated by the predetermined value P or more,
For example, when processing the position correction area shown in FIG. 10, the position correction values of the immediately preceding position correction area are 3, 3 and the position correction values of the hatched position correction area are 5, 5,
If it is more than P (2 in both the Y and X directions), it is determined that the position correction value of the position correction area is an inappropriate value, and the routine proceeds to step 80, where the immediately preceding position correction area (FIG. Substitute the position correction value of

In the above-described processing, a position correction value is given to all the position correction areas, and the given position correction value is written into the position correction value register 66 at every interruption interval, and the master data is stored in the position correction value register 66. The position is corrected for each correction area, and the circuit pattern is inspected by comparison with the defect candidate data.

In the simple inspection according to the second embodiment, since the position is corrected in real time based on the master data, for example, the positioning of a positioning pin (not shown) provided on the inspection table 12 shown in FIG. When the accuracy is low, or the wiring of the circuit pattern becomes thicker or thinner due to a change in conditions such as etching, the position correction cannot be performed with the position correction value set in advance in the first embodiment. Once it is determined, it can be checked again whether a defect actually occurs.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a circuit pattern inspection device according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing processing in a defect matching unit, a master data memory unit, and a data editing unit in FIG.

FIG. 3 is an explanatory diagram showing a configuration of image data taken into the circuit pattern inspection device according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a configuration of a position correction area in FIG.

FIG. 5 is a circuit diagram showing a configuration of a master data XY development line memory and a shift register in FIG. 2;

FIG. 6A is an explanatory diagram of expanded master data, and FIG. 6B is a circuit diagram of a mismatch counting block constituting a mismatch counting circuit.

FIG. 7 is a circuit diagram showing a configuration of a matching circuit with a master data correction function in FIG. 2;

FIG. 8 is a flowchart showing the contents of an area-specific position correction area determining routine.

FIG. 9 is a flowchart showing the contents of a position correction value correction routine.

FIG. 10 is a flowchart showing the contents of a position correction value setting routine.

FIG. 11 is a block diagram illustrating processing in a defect matching unit, a master data memory unit, and a data editing unit according to a simplified process according to the second embodiment of the present invention.

FIG. 12 is an explanatory diagram illustrating a configuration of image data captured at the time of simple processing according to the second embodiment.

13 is a flowchart showing the contents of a position correction value correction routine shown in FIG.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 10 Circuit pattern inspection apparatus 18 Camera 26 Feature extraction part 30 Defect collation part 36 Master data memory part 52 Master data XY expansion line memory + Shift register 54 Mismatch counting circuit 66 Position correction value register 68 Collation circuit with master data correction function

Claims (7)

[Claims] 1. A circuit pattern inspection method that uses one master image data to determine whether a circuit pattern is appropriate in a plurality of circuit pattern inspection apparatuses by comparing the captured image data with each other. (A) A circuit pattern inspection method including a deviation amount measuring step and (B) a propriety judging step: (A) The deviation amount between the master image data and the inspection device, which has the following (1) to (3), is measured. (1) converting the master image data into m (Y direction) × n
(2) The image data of the acquired position correction area is divided into M ×
(3) The developed data and the data obtained by dividing the inspection image data into the position correction areas and comparing them are compared for each of the M × N image processing units. Is determined in the Y and X directions. (B) The master image data is shifted in the Y direction and the X direction for each position correction area by the measured deviation amount, and is compared for each position correction area of the acquired inspection image data to determine whether the circuit pattern is appropriate. Process. 2. The circuit pattern inspection method according to claim 1, wherein the position correction area is obtained by dividing an interrupt interval of a control device for image data taken from a camera into a plurality of sections. 3. The method of determining the amount of deviation in (3) is to count the number of matches or mismatches of M × N image processing units, and determine the amount of shift in the Y and X directions to the position with the highest number of matches. 2. The circuit pattern inspection method according to claim 1, wherein the shift amount is determined as a shift amount for each position correction area. 4. In the determination of the shift amount in (3), when the position having the largest number of matches cannot be specified, or when the number of matches is smaller than a predetermined value, the shift amount of the position correction area is determined as: The circuit pattern inspection method according to any one of claims 1 to 3, wherein a deviation amount of an adjacent position correction area for which the deviation amount can be determined is used. 5. The method according to claim 3, wherein, in determining the shift amount in (3), after determining a position having a large number of coincidences, the determined shift amount is larger than a shift amount of an adjacent position correction area by a predetermined value or more. And determining, when the distance is large, a deviation amount of an adjacent position correction area whose deviation amount can be determined as the deviation amount of the position correction area.
5. The circuit pattern inspection method according to any one of 4. 6. The circuit according to claim 2, wherein the following steps (5) to (8) are performed when it is determined in (B) that the circuit pattern is inappropriate. Pattern inspection method: (5) The master image data is converted into m2 (Y direction) × n
A second position correction area for performing position correction consisting of 2 (X direction) image processing units, and for each second position correction area consisting of one interruption interval of the control device for image data taken in from the camera. Dividing and importing,
(6) a step of developing the captured image data of the second position correction area into M × N images; (7) dividing the expanded data and the inspection image data into a second position correction area and capturing the divided data; Determining the amount of deviation in the Y direction and the X direction for each second position correction area by comparing the data with each other in the M × N image processing units; The image data is converted into Y for each second position correction area.
Determining the suitability of the circuit pattern by comparing the acquired image data for inspection in each of the second position correction areas by shifting in the direction and the X direction. 7. A circuit pattern inspection apparatus which compares master image data with captured inspection image data to determine whether or not a circuit pattern is appropriate: (A) converting the master image data into m (Y direction) × n
(X direction) The image data of the position correction area is divided into each position correction area for performing position correction consisting of image processing units in the (X direction), and the obtained image data of the position correction area is developed into M × N images. For each of the M × N image processing units, data obtained by dividing the image data for position correction into the position correction areas is compared for each of the M × N image processing units, and a value that determines the amount of deviation in the Y direction and the X direction for each position correction area is held. (B) a shift amount holding unit;
The master image data is shifted in the Y direction and the X direction by the measured deviation amount for each position correction area, and is compared for each position correction area of the acquired inspection image data to judge the suitability of the circuit pattern. A circuit pattern inspection apparatus, comprising: