JPS62228103A - Apparatus for measuring substrate - Google Patents

Abstract

PURPOSE:To measure the center coordinates of a pattern with high accuracy even if blur is generated in the original variable density image, by applying the weighting of light and shade to the blurred part of a pattern edge. CONSTITUTION:A control computer 11 inputs the radiation fluoroscopic image transmitted through a substrate as a variable density image from input machinery 31 to store the same in an image memory 32. Thereafter, in an image-to- image operation part 33, the variable density image receives the correction of sensitivity on the basis of the reference image preliminarily stored in the memory 32 and binarized in an image binarizing part 34 to extract a pattern contour line in a logical filter part 35. Next, the operation part 33 calculates the light and shade sum of the variable density image in the vicinity of a contour part and the position vector sum, obtained by applying the weighting of light and shade to said light and shade sum, on the basis of AND between the picture elements of a contour image and the original variable density image at the same position and performs predetermined operation on the basis of the calculation result to make it possible to calculate center coordinates with high accuracy.

Description

Detailed Description of the Invention [Object of the Invention] (Industrial Application Field) The present invention relates to a substrate measurement method for measuring the center of a specific pattern on a substrate, such as the center of a land pattern formed on a substrate or the center of a reference pattern for positioning. Concerning improvements to equipment.

(Prior Art) A conventional substrate measuring device of this type scans the surface of a substrate using an image input means of a TV left camera, and obtains an optical image of a nearby area including a particular pattern on the substrate surface or an optical image of the entire substrate. The image is captured and input as an original grayscale image, and this original grayscale image is binarized using a predetermined threshold value to obtain a binary image, and the center of this binary image is determined and determined as the center position of a specific pattern. It has established.

(Problem to be solved by the invention) However, the board API equipped with the above image processing means
For example, when determining the center of an inner layer pattern or a surface pattern in a multilayer substrate, a fixed image input means inputs an X-ray fluoroscopic image or an optical image with a slight blur as an electrical original grayscale image. Then, the image processing system immediately binarizes the original grayscale image to obtain a binary image and calculates the center position from this binary image, so if the original grayscale image is blurred, the image Errors often occur in the center coordinates of the pattern determined by the processing means.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a substrate measuring device that can measure the central coordinates of a pattern with high precision by weighting the blurred portions near the pattern with shading.

[Structure of the Invention] (Means for Solving the Problems) According to the substrate measuring device according to the present invention, the image input means scans the substrate surface to obtain an original gradation image, and the original gradation image is transferred to the outline extraction means. 2 after binarization or differentiation by
Value is extracted and the contour is extracted, and this contour image and the original grayscale image are logically ANDed to obtain the grayscale sum of the grayscale image near the pattern contour, and this grayscale sum and the position vector sum which is weighted to this grayscale sum are calculated. The center coordinates of the pattern are determined using the method, and the center coordinate data is outputted by the center coordinate data output means.

(Operation) Therefore, by using the above means, the original grayscale image is binarized by the outline extracting means to obtain a binary image, and the outline of the pattern is extracted by the logic between this binary image and the reference image. Extract the outline of the pattern by expanding or emphasizing the pattern edges as necessary, and calculate the sum of shading of the shading image near the pattern by logic between the pixels of the obtained outline image and the original shading image. Since the center coordinates of the pattern are determined from this sum of shading and the sum of the position vectors weighted by the sum of shading, appropriate shading weights are added to the blurred portions of the pattern edges, even if blurring occurs in the original gradation image. The center coordinates of the pattern can be determined with high precision.

(Example) Hereinafter, an example of the present invention will be described with reference to FIGS. 1 to 4. FIG. 1 is a functional block diagram showing the main parts of the device of the present invention, FIG. 2 is a diagram showing an example of the configuration of the device of the present invention, and FIG.
The figure is an operation flowchart of the apparatus of the present invention, and FIG. 4 is a diagram showing a pattern image during image processing. Functionally, the apparatus of the present invention has an image input means A5 which scans the substrate surface to obtain an original gray scale image of the substrate surface or an original gray scale image based on a radioscopic image. Contour extracting means B5 which binarizes or differentiates and then binarizes the grayscale image inputted by , and extracts the outline of the pattern from this binary image and the reference image using a logical filter.
The sum of shading of the shading image in the vicinity of the pattern edge obtained by the logical product between pixels of the contour image and the original gradation image obtained by The center coordinate measuring means C determines the center coordinates of the pattern using a predetermined calculation formula, and the center coordinate data output means outputs the center coordinate data determined by the center coordinate measuring means C.

Next, the configuration of the apparatus of the present invention will be explained in more detail with reference to FIG. In the figure, 11 is a control computer, which sends necessary commands or drive control signals to each part of the device via an input/output bus 12, a direct memory bus 13, and a control line (not shown). It has the function of performing predetermined control and executing image processing based on a predetermined program. Usually, a microprocessor or the like is used as the control computer 11, and is generally called a central processing unit CPU.

For example, a line printer 15, a CRT display section 16, etc. are connected to the input/output bus 12 via an input/output interface 14.

Reference numerals 17 and 18 are a main storage section and an auxiliary storage section that store programs for image processing, necessary fixed constants, predetermined arithmetic expressions, and the like. It also has a direct memory controller (not shown), which connects the main memory section 17 . Auxiliary storage unit 1
8 and an image memory, which will be described later, are directly read out and stored in a direct memory (not shown). Reference numeral 19 denotes an image processing interface having a function of transmitting commands from the control computer 11 to each image processing element to be described later via the multipath 20, or converting them into predetermined signals. An image input device 31 as an image input means A is connected between the multipath 20 derived from the image processing interface 19 and the image data bus 21, and an image memory 32 as a contour extraction means B is connected between the images. The calculation section 332, image binarization section 34, logic filter section 35, etc. are connected.

The image input device 31 uses, for example, a microscope and an optical TV camera, a radiation TV camera, or a radiation line sensor, and captures a substrate surface image or a radiographic image obtained by scanning the substrate and inputs it as an original grayscale image. .

The image memory 32 has a function of storing a result image, etc. during image processing of the original grayscale image inputted by the image input device 31. The inter-image calculation unit 33 performs addition/subtraction or logical product calculation between pixels at the same position of a preset reference image and the gray scale image from the image input device 31 to mutually calculate the mutual relationship between each detection element of the image input device 31. Obtain the sensitivity correction value.

The image binarization unit 34 has a predetermined threshold value set therein, and uses this threshold value to binarize the original grayscale image to obtain a binary image. The logical filter section 35 uses, for example, a contour extraction filter or an expansion logical filter, and performs contour extraction or expansion using these filters.

The center coordinate measuring means C, center coordinate data output means, etc. are configured as part of the image processing system, but these are mainly processed by software by the control computer 11, and the center coordinate data as a result of the processing is processed by the control computer 11. is outputted from a line printer 15 or a CRT display section 16, which are part of the center coordinate data output means.

Next, we will discuss the operation of the device configured as described above in the third section.
This will be explained with reference to the figures and FIG. il) The control computer 11 gives a drive control command to a radiation source (not shown) via the input/output interface 14, for example, based on a program stored in the main storage section 17, the auxiliary storage section 18, etc., and controls the board. While irradiating radiation, controlling an image input device 31 such as a radiation TV camera through an image processing interface 19 and a multipath 20,
Step S: receiving the radiographic image transmitted through the substrate and inputting it as a grayscale image from the image input device 31;
1), storage in the image memory 32 via the image data bus 21; After that, the control computer 11 controls the operation of the inter-image calculation section 33 via the image processing interface 19 as shown in step S2, and then the inter-image calculation section 3
3 performs addition/subtraction or logical product operation between the same pixels of the original gray scale image of the board stored as a table in the image memory 32 and the reference image stored in advance in the image memory 32 etc. to obtain a correction value. Then, this correction value is added to the original grayscale image of the substrate to correct the sensitivity between the detection elements of the TV camera, and the corrected original grayscale image is similarly stored in the image memory 32. Subsequently, the operation of the image binarization unit 34 is controlled based on instructions from the control computer 11, and the original grayscale image, which has been sensitivity-corrected using a predetermined threshold value set in advance, is binarized to obtain a binary image. This image binarization unit 34
If the binarized image 41 has a land, that is, a pattern image 41b in the center of the base image 41a, as shown in FIG. ”, and on the other hand, it is assumed that the pattern image 41b appears as a logical “1” with a tone difference of 255, for example. Therefore, in step S3, based on the binary image 41 obtained as described above, the outline of the land is extracted. This process is carried out by the logical filter section 35.
Then, a pattern outline 41c as shown in FIG. 4(b) is extracted by logic between the binary image 41 and a reference image in which the base image and the pattern image have opposite logical values.

Further, if necessary, in step S4, the
) is expanded n times by the expansion filter of the logic filter section 35 to expand the width of the contour line to 20+1 (FIG. 4C). At this time, the blur width of the pattern edge of the original; light and shade image is 2n+
It should be about 1 pixel. In other words, if the blur width is about 5 pixels, n=2 (n is a natural number).

Next, in step S5, the control computer 11 connects the image processing unit 33 to the image processing interface 19.
is controlled, and the sum of gradations G (i, j) of the gradation images in the vicinity of the outline is obtained by logical product between pixels at the same position of the contour image obtained in step S4 and the original gradation image. Furthermore, in step S6, for the part where G (i, j)≠0, ΣG
The sum of shading of (i, j) and the sum of shading ΣG (i, j) with shading weights, Σ, G (i,
After determining j)・(i, j), if the center coordinates of the 1, J land pattern are (Xc, yc), then the center coordinates of the pattern can be determined using the following equation. After that, the control computer 1 sends the pattern center coordinates (XC, Yc) obtained in step S6 to the line printer 15 or CR7 through the input/output interface 14.
It is output to the display unit 16 (step S7).

Therefore, according to the configuration of the embodiment as described above, the original grayscale image input by the image input means A is binarized using a predetermined threshold value to obtain a binary image, and this binary image and the reference image are After extracting the outline of the pattern using the logic of The center coordinates of the pattern are measured from this sum of shading and the sum of the position vectors, which are weighted by the shading, so even if the original gradation image is blurred, the pattern can be accurately calculated. Center coordinates can be measured. In addition, especially in the case of multilayer substrates, the intensity of the radiographic image is weak, and errors often occur even when measuring the center coordinates of a certain inner layer pattern. However, with this device, shading weights are applied to blurred portions of pattern edges. By attaching this, it is possible to measure the center coordinates of a pattern with much higher precision than in the past.

In addition, in the above embodiment, when extracting the contour, n-1°2.3
...I expanded the binding outline, but for example, n-
If it is considered to be 0, the original grayscale image may be binarized and the contour immediately obtained. Further, the original grayscale image input from the image input means is binarized using one predetermined threshold value, but for example, two predetermined threshold values may be set, and the original grayscale image is binarized using one predetermined threshold value.
It is also possible to extract the outline by converting the pixel into a logic "1" in the case of a pixel that falls between two threshold values, and into a logic "θ" in the case of other pixels. Furthermore, an image in which the edges of the pattern are emphasized is obtained using a spatial product-sum calculation filter that performs first or second differentiation on the original grayscale image input by the image input means, and only the outline is removed using a predetermined threshold value. It may also be configured to binarize to logic "1" and extract the contour.

In addition, the present invention can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] As described in detail above, according to the present invention, it is possible to provide a substrate measuring device that can determine the central coordinates of a pattern with high precision by IJI by weighting the blurred portion of the pattern edge with shading.

[Brief explanation of drawings]

1 to 4 are shown to explain an embodiment of the substrate measuring device according to the present invention. FIG. 1 is a functional block diagram showing the main parts of the device of the present invention, and FIG. FIG. 3 is a diagram explaining the operating procedure of the device, and FIG. 4 is a diagram showing a pattern image obtained during the operation of FIG. 3. A... Image input means, B... Contour extraction means, C...
- Center coordinate measuring means, D... Center coordinate data output means, 11... Control computer, 15... Line printer, 16... CRT display section, 17... Main storage section,
31... Image input device, 32... Image memory, 33
... inter-image calculation section, 34 ... image binarization section, 35.
...Logic filter section, 41... Board image, 41a...
・Vessel image, 41b...Pattern image, 41c...
pattern contour lines. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Figure 2 Figure 3

Claims (5)

[Claims] (1) an image input means for inputting an original grayscale image related to a surface image or an X-ray fluoroscopic image of the substrate obtained by scanning the substrate with an optical image input device or an X-ray image input device;
Contour extraction means that binarizes the original grayscale image input by this image input means at a predetermined threshold value or differentiates it and then binarizes it at a predetermined threshold value, and extracts the outline of a pattern using this binary image. Then, the contour image obtained by this contour extraction means and the original grayscale image are logically ANDed to obtain the grayscale sum of the grayscale image near the pattern contour, and the center coordinates of the pattern are determined from the sum of position vectors weighted to this grayscale sum. What is claimed is: 1. A substrate measuring device comprising: a center coordinate measuring means for determining the center coordinates; and a center coordinate data output means for outputting center coordinate data of a pattern obtained by the center coordinate measuring means. (2) The contour extraction means includes an image binarization unit that binarizes the original grayscale image input by the image input unit using a predetermined threshold; 2. The substrate measuring device according to claim 1, further comprising a logic filter section that extracts the outline of the pattern from the logic of the value image and the reference image. (3) The contour extraction means includes an image binarization unit that binarizes the original grayscale image input by the image input unit using a predetermined threshold; It has a logical filter section that extracts the contour of the pattern from the logic of the value image and the reference image, and a contour expansion means that expands the logical "1" portion obtained by the logical filter section by a predetermined number of pixels. A substrate measuring device according to claim 1. (4) The contour extracting means has two predetermined threshold values set in advance, and selects a pixel that falls between the two predetermined threshold values in the original grayscale image input by the image input means with a logic "1". ”, and other pixels are respectively binarized to logic “0”, and the outline of the pattern is extracted by logic between this binary image and a reference image. Device. (5) The contour extraction means includes an edge emphasis means for differentiating the original grayscale image input by the image input means using a spatial product-sum operation filter to emphasize edges, and an edge emphasis means for applying edge emphasis to a predetermined image by the edge emphasis means. 2. The substrate measuring device according to claim 1, further comprising an image binarization section that binarizes the image using a threshold value of .