A method of reticle strip registration

Abstract

The application discloses a mask strip registration method, and the steps of the method comprise the following steps: establishing a sub-strip registration reference point array; registering a Mask image divided on a sub-strip and a mask plate design file generated graph, and sequentially sliding the Mask image by a pixel step to cover a corresponding size local image on the mask plate design file generated graph; calculating the difference of each pixel point of the Mask image and the corresponding local image after each sliding, presetting a rough threshold, when the absolute value of the difference after the calculation is smaller than the rough threshold, assigning 0 to the difference, then accumulating and summing the difference to obtain a rough difference sum, and saving the rough difference sum to a corresponding rough registration result matrix A; assigning a mark value to the position with the difference sum of 0 in the rough registration result matrix A, and using the registration result of the corresponding sub-strip in the registration reference point array of the previous strip for registration when registering adjacent strips; and the method can reduce the registration range, shorten the registration time, and improve the registration efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to a mask strip registration method, belonging to the technical field of semiconductor detection. BACKGROUND

[0002] At present, in the mask defect detection, the whole scanning strip is cut into multiple sub-strips according to the cutting strategy and fixed size, and the sub-strips are arranged in matrix and indexed in the column direction. The mask array image has translational symmetry, and when one pattern or a group of patterns is offset by a distance in a certain direction, it can coincide with another pattern or a group of patterns. The array pattern has periodic arrangement characteristics, that is, a pattern or a group of patterns repeatedly appears in the image area. The most common case of array pattern in the mask plate is CNT layer, which can be used as a conductive path in integrated circuits or as a channel of field effect transistor. The pattern in the mask plate is usually presented in the form of array pattern. Since the overall image generated by GDS (including but not limited to design files in formats such as GDS and OASIS) is larger than the Mask image screen obtained by the camera, when the pattern is registered, because the pattern has the array arrangement characteristic, the reference points for registration can be obtained at multiple positions. While avoiding the error of the array pattern registration point reference position, the found position is a local optimal registration point, which is the key to ensure the accuracy and speed of registration and detection. SUMMARY

[0003] The technical problem to be solved by the present application is to overcome the defects of the prior art and provide a mask strip registration method which can reduce the registration range, shorten the registration time and improve the registration efficiency.

[0004] To solve the above technical problems, the technical scheme of the present application is as follows: a mask strip registration method, the steps of the method comprising:

[0005] S01: establishing a sub-strip registration reference point array and initializing the registration reference point position of each sub-strip;

[0006] S02: registering the Mask image divided on the sub-strip and the mask design file generated image, and sliding the Mask image by pixel steps in turn to cover the corresponding size of the local image on the mask design file generated image; wherein, after each sliding, the difference between each pixel point of the Mask image and the corresponding local image is calculated, a preset rough threshold is set, when the absolute value of the difference after calculation is less than the rough threshold, the difference at this position is assigned 0, then the difference is accumulated to obtain a rough difference sum, and the rough difference sum is saved in the corresponding rough registration result matrix A;

[0007] S03: In the rough registration result matrix A, mark the position where the difference sum is 0; when no marked value appears in the rough registration result matrix A, the position corresponding to the minimum rough difference sum in the rough registration result matrix A is taken as the registration reference point; when a marked value exists in the rough registration result matrix A, the position corresponding to the marked value is taken as the registration reference point.

[0008] Further, step S02 further includes: presetting a fine threshold value; when the difference value after the difference is less than the fine threshold value, the difference value is processed by assigning 0 to the difference value, and then the difference value is accumulated and summed to obtain a fine difference sum, and saved to the corresponding fine registration result matrix B;

[0009] The method further includes the following steps:

[0010] S04: Mark the adjacent connected domain with multiple registration reference points as a region marker, and count the number of region marker positions in the corresponding region of the substrip that meet the requirements: if there are multiple region markers in the region, it is considered that the registration reference point position marked by the substrip position cannot be referenced by other strips, and if there is only one region marker in the region, the position corresponding to the region marker needs to be saved to the substrip registration reference point array;

[0011] S05: The position of the registration reference point obtained based on the rough registration result matrix A is compared with the corresponding position in the fine registration result matrix B, and the minimum fine difference sum in the fine registration result matrix B is found in the region range of the corresponding position. After all the minimum fine difference sums are found, the corresponding fine registration reference point position is obtained;

[0012] S06: The fine registration reference point position obtained in step S05 is referenced according to the reference requirement specified in step S04 and saved to the substrip registration reference point array.

[0013] Further, the method further includes the following steps:

[0014] S07: After the registration of each column of substrips is completed, the array that cannot be referenced in the entire column of substrip registration reference point arrays is fitted with the registration reference point position of the array that can be referenced, thereby obtaining the registration reference point at this position.

[0015] Further, the Mask image is sequentially stepped and slid in pixels, specifically:

[0016] a: The Mask image is sequentially stepped and slid in pixels in the row direction at the start position of the row, and after the row direction sliding is completed, the Mask image is returned to the start position of the row and then stepped and slid down by one pixel in the column direction;

[0017] b: Step a is performed in a new row;

[0018] c: sequentially perform step b until the last row sliding ends.

[0019] After the above technical scheme is adopted, the registration reference point is provided by calculation for subsequent sub-bands to be used for registration reference, the registration range is reduced and the registration time can be shortened, and the method is used to judge whether the current registration reference point position can be used for reference by other bands. The reference registration point is used to divide the local area of the design file generated graph for registration, which can speed up the registration speed; at the same time, the local optimal point can be avoided as the registration result output in the registration process, and multiple local optimal positions can be counted as alternative positions in rough registration, and then the global optimal registration point in the alternative local optimal points can be found by means of the registration reference point twice registration, so that the defects can be better identified when the design file generated graph and the mask obtained by the camera are processed. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The mask image of the mask strip registration method of the application;

[0021] Figure 2 The design file generated graph of the mask of the application;

[0022] Figure 3 The mask image and the design file generated graph of the mask of the application; Figure 1

[0023] Figure 4 The mask image and the design file generated graph of the mask of the application; Figure 2 DETAILED DESCRIPTION

[0024] In order to make the content of the application more easily and clearly understood, the application will be further described in detail below according to specific embodiments and in combination with the drawings.

[0025] As shown in Figures 1-4 A mask strip registration method, the steps of the method include:

[0026] S01: Establish a sub-band registration reference point array, initialize the registration reference point position of each sub-band, and record the registration point position of each column of sub-bands of the scanned strip to the corresponding sub-band registration reference point array.

[0027] ​​S02: The Mask image divided on the sub-stripes and the mask design file generation image are registered, and the Mask image is slid by pixel step by step to cover the corresponding size of the local image on the mask design file generation image; wherein, after each sliding, the difference of each pixel point of the Mask image and the corresponding local image is calculated, a preset rough threshold is set, when the absolute value of the difference after calculation is less than the rough threshold, the difference at this position is assigned 0, then the difference is accumulated to obtain a rough difference sum, and saved to the corresponding rough registration result matrix A;

[0028] S03: In the rough registration result matrix A, the position of the difference sum of 0 is assigned a mark value; when there is no mark value in the rough registration result matrix A, the position of the minimum rough difference sum in the rough registration result matrix A is taken as the registration reference point; when there is a mark value in the rough registration result matrix A, the position corresponding to the mark value is taken as the registration reference point.

[0029] In the embodiment, in the design, the whole area of the mask design file generation image is not completely taken as the registration area, but a local area is selected as the registration area of the Mask image (this local area is also larger than the Mask image), but the size of the whole rough registration result matrix A is designed according to the whole area of the mask design file generation image, and the data is all set to 0 in the initialization;

[0030] Specifically, step S02 further includes: a preset fine threshold is set, when the absolute value of the difference after calculation is less than the fine threshold, the difference at this position is assigned 0, then the difference is accumulated to obtain a fine difference sum, and saved to the corresponding fine registration result matrix B;

[0031] The method further includes the following steps:

[0032] S04: The adjacent connected domain with multiple registration reference points is marked as a region marker, and the number of region marker meeting the requirements in the corresponding region of the sub-stripes is counted: if there are multiple region markers in the region, it is considered that the registration reference point position marked by this sub-stripes cannot be referenced by other strip, if there is only one region marker in the region, the position corresponding to the region marker needs to be saved to the sub-stripes registration reference point array;

[0033] S05: The position of the registration reference point obtained based on the rough registration result matrix A is compared with the corresponding position in the fine registration result matrix B, and the minimum fine difference sum in the fine registration result matrix B is found in the region range of the corresponding position, after all the minimum fine difference sums are found, the corresponding fine registration reference point position is obtained;

[0034] S06: The fine registration reference point obtained in step S05 is referenced according to the reference requirement specified in step S04, and is saved into the sub-strip registration reference point array.

[0035] Specifically, the method further comprises the steps of:

[0036] S07: After the registration of each column of sub-strips is completed, the registration reference points in the sub-strip registration reference point array of the entire column that cannot be referenced are fitted with the registration reference point positions of the array that can be referenced, so as to obtain the registration reference points here.

[0037] In this embodiment, the adjacent sub-strips are calculated according to the positions (with the addition of the displacement in the row direction) and are provided for use in the registration of the adjacent sub-strips. The reference point relative positions of the pattern design file are generated by using the registration reference point positions, the available registration area is divided inwardly and reduced, and thus the operation is reduced and the registration speed is improved.

[0038] In summary, by using the method, the registration position information of the sub-strips in the adjacent column can be provided for the registration position reference of the adjacent sub-strips in the subsequent column, the registration division area can be reduced, and the registration time length can be shortened. At the same time, when the optimal registration reference point is screened, instead of directly finding the minimum value (i.e., the minimum sum of differences) as the registration point, a wide range of search is performed by using a rough threshold, a range of the area that meets the requirement is found, the difference sum obtained by using a fine threshold is further calculated in each registration reference point area range, and thus the final registration result is compared, which can effectively avoid the problem of falling into local optimization of registration when multiple positions of the array image are matched.

[0039] Specifically, as shown in Figure 3 , 4 , the Mask image is sequentially stepped and slid by pixels, specifically as follows:

[0040] a: The Mask image is sequentially stepped and slid by pixels in the row direction at the start position of the row, after the sliding in the row direction is completed, the Mask image is returned to the start position of the row and then is stepped and slid by one pixel downward in the column direction;

[0041] b: Step a is performed in a new row;

[0042] c: Step b is sequentially performed until the sliding in the last row is completed.

[0043] The above-described specific embodiments further specifically describe the technical problems solved by the present application, the technical solutions and the beneficial effects. It should be understood that the above-described embodiments are merely specific embodiments of the present application and are not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A reticle strip registration method, characterized by, The steps of the method comprise: S01: establishing a sub-strip registration reference point array, and initializing the registration reference point position of each sub-strip; S02: registering the mask image divided on the sub-strip and the mask plate design file generated image, and sequentially sliding the mask image by pixel steps to cover the local image of corresponding size on the mask plate design file generated image; wherein, after each sliding, the difference of each pixel point of the mask image and the corresponding local image is calculated, a preset rough threshold is set, when the absolute value of the difference after calculation is less than the rough threshold, the difference at this position is assigned 0, then the difference is accumulated to obtain a rough difference sum, and the rough difference sum is saved in the corresponding rough registration result matrix A; S03: in the rough registration result matrix A, the position with a difference sum of 0 is assigned a mark value; when no mark value appears in the rough registration result matrix A, the position with the minimum rough difference sum in the rough registration result matrix A is taken as the registration reference point; when the mark value exists in the rough registration result matrix A, the position corresponding to the mark value is taken as the registration reference point; In step S02, a preset fine threshold is further included, when the absolute value of the difference after calculation is less than the fine threshold, the difference at this position is assigned 0, then the difference is accumulated to obtain a fine difference sum, and the fine difference sum is saved in the corresponding fine registration result matrix B; The method further comprises the following steps: S04: taking the connected domain adjacent to the mark value as a region mark bit, and counting the number of region mark bits in the corresponding region of the sub-strip that meet the requirements: if there are multiple region mark bits in the region, it is considered that the registration reference point position marked by the sub-strip cannot be referenced by other strips; if there is only one region mark bit in the region, the position corresponding to the region mark bit needs to be saved in the sub-strip registration reference point array; S05: comparing the position of the registration reference point obtained based on the rough registration result matrix A with the corresponding position in the fine registration result matrix B, and finding the minimum fine difference sum in the region range of the corresponding position in the fine registration result matrix B, after finding all the minimum fine difference sums, the corresponding fine registration reference point position is obtained; S06: referring to the fine registration reference point position obtained in step S05 according to the reference requirements specified in step S04, and saving it in the sub-strip registration reference point array.

2. The reticle strip registration method of claim 1, wherein: Further comprising the step: S07: after the registration of each column of sub-strips is completed, the array that cannot be referenced in the registration reference point array of all sub-strips in the column is fitted with the registration reference point position of the array that can be referenced, so as to obtain the registration reference point at this position.

3. The reticle strip registration method of claim 1, wherein: The sequential pixel step sliding of the mask image is specifically: a: sliding the mask image by pixel steps in the row direction at the starting position of the row, after the sliding in the row direction is completed, returning to the starting position of the row and then sliding the mask image by one pixel step downward in the column direction; b: performing step a in a new row; c: sequentially performing step b until the sliding of the last row is completed.

Images