A method for automatically drawing a region of interest and related products

Abstract

The application discloses a method for automatically drawing a region of interest and a related product, which can be applied to the technical field of semiconductors. The method comprises the following steps: acquiring a full-die image of a die to be drawn based on a preset optical mode; determining a target gray scale corresponding to each pixel point in the full-die image based on a preset gray scale range; grouping each pixel point according to the target gray scale and obtaining a first group set; determining a target group from the first group set based on a preset rule, and performing region of interest drawing based on the pixel points corresponding to the target group. In this way, the application does not depend on a wafer design layout, but only distinguishes the types of the region of interest through a gray scale difference, thereby realizing automatic drawing of the region of interest and improving the flexibility of region of interest drawing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor technology, and in particular, to a method for automatically drawing a region of interest and related products. BACKGROUND

[0002] A region of interest (ROI) refers to a part of a region in an image (or a video frame) that needs to be paid special attention to, needs to be further analyzed or processed.

[0003] It is tedious and difficult to configure a detection recipe ROI on an actual production line. Because with the development of semiconductor technology, the die on an advanced process wafer is getting smaller and smaller, and the pattern is getting more and more complex, the number of ROIs can reach tens of thousands, or even millions. The existing method for configuring an ROI is usually to automatically draw by analyzing the design layout (GDS) of the wafer. That is, the existing method for drawing an ROI cannot automatically draw without relying on the implementation of GDS, thereby causing the existing method for drawing an ROI to have a low flexibility problem.

[0004] Therefore, how to improve the flexibility of the drawing of the region of interest is a problem that those skilled in the art urgently need to solve. SUMMARY

[0005] Based on the above problems, the present application provides a method for automatically drawing a region of interest and related products, which does not rely on a wafer design layout, and only distinguishes the types of the region of interest through a gray scale difference, thereby realizing automatic drawing of the region of interest and improving the flexibility of the drawing of the region of interest.

[0006] In a first aspect, an embodiment of the present application provides a method for automatically drawing a region of interest, comprising:

[0007] obtaining a full-die image of a to-be-drawn die based on a preset optical mode;

[0008] determining a target gray scale corresponding to each pixel point in the full-die image based on a preset gray scale range;

[0009] grouping each pixel point according to the target gray scale, and obtaining a first group set;

[0010] determining a target group from the first group set based on a preset rule, and performing region of interest drawing based on the pixel points corresponding to the target group.

[0011] Optionally, the obtaining of the full-die image of the to-be-drawn die based on the preset optical mode comprises:

[0012] acquire a surface image of a wafer to be measured, in which the die to be drawn is located, based on a preset optical mode;

[0013] collect a full-die image corresponding to the die to be drawn from the surface image based on a line scanning technique.

[0014] Optionally, the method further comprises:

[0015] select N optical modes from a set of optical modes as the preset optical modes; the N is an integer greater than or equal to 1;

[0016] any two optical modes in the set of optical modes are different in at least one of a wavelength band, a polarization, and a focusing mode.

[0017] Optionally, the collecting a full-die image corresponding to the die to be drawn from the surface image based on a line scanning technique comprises:

[0018] select K target dies from the surface image; the K is an integer greater than or equal to 1; the target dies are the same type of die as the die to be drawn;

[0019] collect a full-die image corresponding to the target dies from the surface image based on a line scanning technique.

[0020] Optionally, the selecting K target dies from the surface image comprises:

[0021] select K target dies from a middle region of the surface image;

[0022] when the K is greater than 1, the selected target dies are adjacent.

[0023] Optionally, the method further comprises:

[0024] when the K is greater than or equal to 2, perform sub-pixel alignment processing on the full-die images corresponding to the target dies, so that the number of pixels in each full-die image is the same and completely aligned;

[0025] merge all the full-die images collected under the same optical mode; the gray value of each pixel point in the merged full-die image is the median of the gray values of the corresponding pixel points in the full-die images before merging;

[0026] perform image noise reduction processing on the merged full-die image.

[0027] Optionally, the pixel points corresponding to each group in the first group set correspond to only one target gray scale under each optical mode.

[0028] Optionally, the determining a target group from the first group set based on a preset rule comprises:

[0029] determining the number of pixel points corresponding to each group in the first group set;

[0030] sorting the groups according to the number of pixel points from more to less;

[0031] taking the groups ranked in the front X as target groups; the X is an integer greater than or equal to 1.

[0032] Optionally, the region of interest drawing based on the pixel points corresponding to the target groups comprises:

[0033] labeling all pixel points in the target groups in a preset pixel image;

[0034] eliminating isolated pixel points from the all pixel points, and constructing an external polygon based on the remaining pixel points;

[0035] splitting the external polygon into rectangles, and performing region of interest drawing based on the rectangles.

[0036] Optionally, the labeling all pixel points in the target groups in a preset pixel image comprises:

[0037] reserving pixel values of all pixel points in the target groups in the preset pixel image, and setting pixel values of pixel points at positions other than the target groups as empty.

[0038] In a second aspect, an embodiment of the present application provides a device for automatically drawing a region of interest, comprising:

[0039] an acquisition module configured to acquire a full die image of a die to be drawn based on a preset optical mode;

[0040] a determination module configured to determine a target gray scale corresponding to each pixel point in the full die image based on a preset gray scale range;

[0041] a grouping module configured to group each pixel point according to the target gray scale, and obtain a first group set;

[0042] a drawing module configured to determine a target group from the first group set based on a preset rule, and perform region of interest drawing based on pixel points corresponding to the target group.

[0043] In a third aspect, an embodiment of the present application provides a device for automatically drawing a region of interest, comprising:

[0044] a memory configured to store a computer program;

[0045] a processor configured to implement the steps of the method for automatically drawing a region of interest when the computer program is executed.

[0046] In a fourth aspect, an embodiment of the present application provides a readable storage medium, and the readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the method for automatically drawing a region of interest are implemented.

[0047] From the above technical solutions, compared with the prior art, the present application has the following advantages:

[0048] The method for automatically drawing a region of interest provided by the present application comprises the following steps: first, based on a preset optical mode, a full-die image of a die to be drawn is obtained. Then, based on a preset gray scale range, a target gray scale corresponding to each pixel point in the full-die image is determined, each pixel point is grouped according to the target gray scale, and a first group set is obtained. Finally, a target group is determined from the first group set based on a preset rule, and a region of interest is drawn based on the pixel points corresponding to the target group. In this way, the present application does not rely on a wafer design layout, but only distinguishes the types of regions of interest through gray scale differences, thereby realizing automatic drawing of a region of interest and improving the flexibility of region of interest drawing. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 A flowchart of the method for automatically drawing a region of interest provided by an embodiment of the present application is shown in the figure;

[0050] Figure 2 A schematic diagram of an optical mode set provided by an embodiment of the present application is shown in the figure;

[0051] Figure 3 A schematic diagram of selecting a target die provided by an embodiment of the present application is shown in the figure;

[0052] Figure 4 A structural schematic diagram of a full-die image provided by an embodiment of the present application is shown in the figure;

[0053] Figure 5 A flowchart of image data processing provided by an embodiment of the present application is shown in the figure;

[0054] Figure 6 A schematic diagram of pixel grouping and statistics provided by an embodiment of the present application is shown in the figure;

[0055] Figure 7 A flowchart of region of interest drawing provided by an embodiment of the present application is shown in the figure;

[0056] Figure 8 A structural schematic diagram of a device for automatically drawing a region of interest provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0057] As described above, the existing method of drawing the ROI has the problem of low flexibility. Specifically, the work of configuring the detection recipe ROI in the actual production line is tedious and difficult. Because with the development of semiconductor technology, the die on the advanced process wafer is getting smaller and smaller, and the pattern is getting more and more complex, the number of ROIs can reach tens of thousands, even hundreds of thousands or millions. And the existing method of configuring the ROI has two kinds, one is to automatically draw by analyzing the design layout of the wafer, and the other is to manually draw by artificial according to the light and shade gray scale difference of different regions of the image. Among them, the manual drawing efficiency is too low, and the former cannot realize automatic drawing without relying on GDS, which leads to the problem of low flexibility of the existing method of drawing the ROI.

[0058] To solve the above problems, the embodiment of the present application provides a method for automatically drawing a region of interest, which comprises the following steps: first, based on a preset optical mode, obtaining a full-die image of a die to be drawn. Then, based on a preset gray scale range, determining the target gray scale corresponding to each pixel point in the full-die image, and grouping each pixel point according to the target gray scale to obtain a first group set. Finally, determining a target group from the first group set based on a preset rule, and performing region of interest drawing based on the pixel points corresponding to the target group.

[0059] In this way, the present application does not rely on the wafer design layout, but only distinguishes the types of regions of interest by gray scale difference, thereby realizing automatic drawing of the region of interest and improving the flexibility of the region of interest drawing.

[0060] It should be noted that the method for automatically drawing a region of interest and related products provided by the embodiment of the present application can be applied to the field of semiconductor technology. The above is only an example and does not limit the application of the method for automatically drawing a region of interest and related products provided by the present application.

[0061] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0062] Figure 1 The flowchart of the method for automatically drawing a region of interest provided by the embodiment of the present application is shown. In combination with the method for automatically drawing a region of interest provided by the embodiment of the present application, the method can include: Figure 1

[0063] S101: Based on a preset optical mode, obtaining a full-die image of a die to be drawn. ​

[0064] In practical applications, the preset optical mode refers to a pre-selected stable propagation form or resonance form of light energy in space when collecting images of the wafer to be measured. It can be understood that the images of the wafer to be measured collected under different optical modes are generally different. The die to be drawn refers to a certain type of die in the region of interest that needs to be drawn on the wafer to be measured. In general, the type of die on the same wafer is consistent. The full die image is the complete die image of the die to be drawn, that is, the Full Die image. In the embodiments of the present application, the full die image corresponding to the die to be drawn on the wafer to be measured needs to be obtained under the preset optical mode. It should be noted that the method provided by the present application needs to image the surface of the wafer when obtaining the full die image, so it is mainly applied to bright field detection, that is, the method provided by the embodiments of the present application is mainly for the drawing of the region of interest in the detection recipe creation process of the bright field defect detection equipment. In addition, it should be noted that the method provided by the present application is not limited to bright field defect detection equipment.

[0065] In addition, since the selection of the preset optical mode is not the same, the embodiments of the present application can be described with respect to one possible selection method.

[0066] In one case, the method further comprises:

[0067] selecting N optical modes from the set of optical modes as the preset optical mode; N is an integer greater than or equal to 1;

[0068] There is at least any one difference between any two optical modes in the set of optical modes, such as wavelength, polarization and focusing mode.

[0069] In practical applications, there is at least any one difference between different optical modes, such as wavelength, polarization and focusing mode. Figure 2 A schematic diagram of a set of optical modes provided by the embodiments of the present application is shown. In combination with Figure 2 As shown, different optical modes correspond to different mode numbers, such as mode 1 corresponding to wavelength A bright field mode, mode 2 corresponding to wavelength B bright field mode, mode 3 corresponding to wavelength A dark field mode, etc. The preset optical mode can be selected from the set of optical modes, and the number of selections can be any integer greater than or equal to 1. It can be understood that the present application determines which regions are indeed the same type of ROI by whether the optical response of the selected optical mode by the pixel points is the same. Therefore, if the optical modes in the preset optical mode are selected too few, it may not be possible to finally distinguish different types of ROIs.

[0070] In addition, since the way of obtaining the full die image is not the same, the embodiments of the present application can be described with respect to one possible acquisition method. In practical applications, the preset optical mode refers to a pre-selected stable propagation form or resonance form of light energy in space when collecting images of the wafer to be measured. It can be understood that the images of the wafer to be measured collected under different optical modes are generally different. The die to be drawn refers to a certain type of die in the region of interest that needs to be drawn on the wafer to be measured. In general, the type of die on the same wafer is consistent. The full die image is the complete die image of the die to be drawn, that is, the Full Die image. In the embodiments of the present application, the full die image corresponding to the die to be drawn on the wafer to be measured needs to be obtained under the preset optical mode. It should be noted that the method provided by the present application needs to image the surface of the wafer when obtaining the full die image, so it is mainly applied to bright field detection, that is, the method provided by the embodiments of the present application is mainly for the drawing of the region of interest in the detection recipe creation process of the bright field defect detection equipment. In addition, it should be noted that the method provided by the present application is not limited to bright field defect detection equipment.

[0071] In one case, the S101: acquiring a full die image of a die to be drawn based on a preset optical mode, can specifically include:

[0072] Acquiring a surface image of a wafer to be measured where the die to be drawn is located based on a preset optical mode;

[0073] Collecting a full die image corresponding to the die to be drawn from the surface image based on line scanning technology.

[0074] In actual application, in combination with the above, the die to be drawn located on the wafer to be measured refers to a type of die, and generally only one type of die exists on a wafer. If the preset optical mode only includes one optical mode, the surface image of the wafer to be measured where the die to be drawn is located is acquired under the optical mode first. If the preset optical mode includes multiple optical modes (two or more), the surface image of the wafer to be measured where the die to be drawn is located needs to be acquired under each optical mode. The surface image is generally acquired by a high-precision optical microscopic imaging system. Further, in combination with a high-steady-state synchronous scanning system and a high-sensitivity line array camera (Time Delay Integration, TDI), a complete die image (i.e., a full die image) corresponding to the die to be drawn is collected from the surface image based on line scanning technology.

[0075] In addition, since the way of collecting the full die image corresponding to the die to be drawn from the surface image is not the same, the embodiments of the present application can describe one possible collection method.

[0076] In one case, the S101: acquiring a full die image of a die to be drawn based on a preset optical mode, can specifically include:

[0077] Selecting K target dies from the surface image; the K is an integer greater than or equal to 1; the target dies are the same type of die as the die to be drawn;

[0078] Collecting full die images corresponding to the target dies from the surface image based on line scanning technology.

[0079] Further, the selecting K target dies from the surface image includes:

[0080] Selecting K target dies from the middle region of the surface image;

[0081] When K is greater than 1, each target die selected is adjacent.

[0082] In practical applications, in order to avoid the influence of various accidental errors in the collection process on the results, a plurality of dies can be collected. It can be understood that on the wafer to be measured, any die belonging to the type of die to be drawn can be taken as a target die, and the full die image corresponding thereto can be used for subsequent drawing of the region of interest. However, in order to reduce errors, the number of samples can be increased, and the more optimal samples can be selected for processing. Specifically, Figure 3 A schematic diagram of selecting a target die is provided for an embodiment of the present application. As shown in Figure 3 The center of the wafer (process stability) and the multiple dies in the adjacent position can be selected as target dies. Figure 3 In the above embodiment, three dies in the adjacent position are selected as target dies (i.e. gray filled blocks). Further, after selecting the target dies, the complete die images (i.e. full die images) corresponding to the selected target dies are collected from the surface image based on the line scanning technology.

[0083] Figure 4 A structure diagram of a full die image is provided for an embodiment of the present application. As shown in Figure 4 The three dies are selected as target dies, for example, and one full die image corresponding to each of the three target dies under mode 1, one full die image corresponding to each of the three target dies under mode 2, and one full die image corresponding to each of the three target dies under mode 3 are obtained after collection. Under the three modes, the bright and dark field distribution on the full die image corresponding to the same target die can be different.

[0084] S102: Determine the target gray scale corresponding to each pixel point in the full die image based on a preset gray scale range.

[0085] In practical applications, there are many pixel points in a full grain image, and the gray values of each pixel point are not the same. In view of so many possibilities, the present application needs to compress the gray levels of each full grain image, and merge the gray levels of the images to reduce the bit depth of the image. Specifically, the gray value range is [0, 255], and the present application embodiment can divide the range into X parts, for example, X = 3, the divided gray value range can be [0, 84], [85, 170] and [171, 255], and the corresponding gray scale can be set as: [0, 84] corresponds to gray scale 1, [85, 170] corresponds to gray scale 2, and [171, 255] corresponds to gray scale 3. It can be understood that other parts, other gray value ranges and corresponding gray scales can also be set. In combination with the above example, the divided gray value range [0, 84], [85, 170] and [171, 255] is the preset gray scale range, further, the gray value corresponding to each pixel point in the full grain image is determined, then the pixel points falling into [0, 84] are corresponding to gray scale 1, that is, gray scale 1 is the target gray scale corresponding to this kind of pixel points; the pixel points falling into [85, 170] are corresponding to gray scale 2, that is, gray scale 2 is the target gray scale corresponding to this kind of pixel points; the pixel points falling into [171, 255] are corresponding to gray scale 3, that is, gray scale 3 is the target gray scale corresponding to this kind of pixel points. For different gray scales, a gray value can be selected to represent the entire gray scale (generally the median of the gray value range corresponding to each gray scale), and there are at most three selected gray values in the full grain image after compressing the gray levels.

[0086] In addition, since the number of target grains is different, the corresponding image processing method is different, therefore, the present application embodiment can illustrate a possible processing method.

[0087] In one case, the method further comprises:

[0088] When the K is greater than or equal to 2, the full grain images corresponding to each target grain are subjected to sub-pixel alignment processing, so that the number of pixels in each full grain image is the same and completely aligned;

[0089] Merge all full grain images collected under the same optical mode; the gray value of each pixel point in the merged full grain image is the median of the gray value of the corresponding each pixel point in the full grain image before merging;

[0090] The merged full grain image is subjected to image noise reduction processing.

[0091] In practical applications, if the number of target grains is 1, the full grain image corresponding to the target grain is directly subjected to image noise reduction processing, and then the gray level of the full grain image is compressed. When the number of target grains is greater than or equal to 2, the full grain images collected for different target grains under the same optical mode need to be merged. Specifically, Figure 5 A flowchart of image data processing is provided for the embodiments of the present application. As shown in Figure 5 three different optical modes, and three target grains are selected as examples, three full grain images can be collected under mode 1, mode 2 and mode 3, respectively. Then, the full grain images corresponding to all target grains under all optical modes are subjected to sub-pixel alignment processing (image alignment) to ensure that the number of pixels of all collected images is the same and completely aligned, and the gray values can be directly compared pixel by pixel. Further, the median gray values of the pixels at the same position in the multiple full grain images under the same optical mode are taken pixel by pixel to achieve image merging. For example, the gray values of the pixels at the same position in the three full grain images are 40, 50 and 60, respectively. After merging the three full grain images, the gray value of the pixel at the same position in the new full grain image obtained is 50. Then, the merged full grain image can be subjected to image noise reduction processing. Further, the full grain image after noise reduction can be subjected to gray level compression processing.

[0092] S103: Grouping each pixel point according to the target gray level, and obtaining a first group set.

[0093] Further, the pixel points corresponding to each group in the first group set correspond to the same target gray level under each optical mode.

[0094] In practical applications, the pixel points corresponding to the same target gray level under each optical mode need to be grouped in the same group, and then all obtained groups are sorted to obtain the first group set. It should be noted that the same target gray level here refers to the same target gray level under the same optical mode. Specifically, Figure 6 A schematic diagram of pixel grouping and statistics is provided for the embodiments of the present application. As shown in Figure 6As shown, it is assumed that the preset optical modes include three different optical modes (mode 1, mode 2 and mode 3), and three target gray scales (gray scale 1, gray scale 2 and gray scale 3) are set as an example, and based on the grouping rule, in the same gray scale corresponding to the pixel points in one optical mode, if there are pixel points still corresponding to the same gray scale in each mode, such pixel points are classified into the same group. In other words, the pixel points in group 1 correspond to the same gray scale (gray scale 1) in mode 1, the same gray scale (gray scale 1) in mode 2 and the same gray scale (gray scale 2) in mode 3; the pixel points in group 2 correspond to gray scale 1 in mode 1, gray scale 1 in mode 2 and gray scale 2 in mode 3; the pixel points in group 3 correspond to gray scale 1 in mode 1, gray scale 1 in mode 2 and gray scale 3 in mode 3; the pixel points in group 4 correspond to gray scale 1 in mode 1, gray scale 2 in mode 2 and gray scale 1 in mode 3; the pixel points in group 5 correspond to gray scale 1 in mode 1, gray scale 2 in mode 2 and gray scale 2 in mode 3; the pixel points in group 6 correspond to gray scale 1 in mode 1, gray scale 2 in mode 2 and gray scale 3 in mode 3, and so on. It can be understood that if the preset optical modes include N different optical modes, and M target gray scales are set, there are at most M N groups.

[0095] S104: Determine a target group from the first group set based on a preset rule, and perform region of interest drawing based on the pixel points corresponding to the target group.

[0096] In actual application, the pixel points in each group are considered as the same type, and all the pixel points corresponding to each type can constitute a preselected region, and the region of interest is selected from the preselected regions. Specifically, the present embodiment introduces a preset rule, selects a target group from the first group set based on the preset rule, and then draws the region of interest based on the region constituted by the pixel points in the target group. The preset rule can be a pixel point number threshold, for example, the pixel point number threshold is set to 50, and the group containing more than 50 pixel points is recorded as the target group, and then the region of interest is drawn based on the region constituted by the pixel points in the selected target group.

[0097] In addition, since the way of determining the target group is not the same, the present embodiment can explain one possible determination method.

[0098] In one case, the target group is determined from the first group set based on a preset rule, including:

[0099] Determine the number of pixel points corresponding to each group in the first group set;

[0100] sort the groups according to the pixel point quantity from more to less;

[0101] select the first X groups as the target groups, where X is an integer greater than or equal to 1.

[0102] In practical applications, the preset rule can also be set to select the target groups according to the pixel point quantity. Specifically, assuming that nine groups are obtained. Further, if the preset rule is to select the first X groups (assuming X is equal to 3) with the most pixel points, first determine how many pixel points each of the nine groups corresponds to, and then sort the groups according to the number of corresponding pixel points from more to less. Finally, combine the sorting results to take the first three groups as the target groups. For example, if the pixel points corresponding to group 1 to group 9 are 70, 32, 26, 10, 5, 90, 80, 55, and 81, respectively, sort the groups according to the number of corresponding pixel points, and the sorting result is: group 6, group 9, group 7, group 1, group 8, group 2, group 3, group 4, and group 5. According to the preset rule, the selected target groups are group 6, group 9, and group 7. Of course, X can also take other suitable positive integer values, that is, X is an integer greater than or equal to 1.

[0103] In addition, since the ways of drawing the region of interest based on the pixel points corresponding to the target groups are not the same, the embodiments of the present application can explain one possible drawing method.

[0104] In one case, the drawing of the region of interest based on the pixel points corresponding to the target groups comprises:

[0105] label all pixel points in the target groups in a preset pixel image;

[0106] remove isolated pixel points from the all pixel points, and construct a circumscribed polygon based on the remaining pixel points;

[0107] split the circumscribed polygon into rectangles, and draw the region of interest based on the rectangles.

[0108] Further, the labeling of all pixel points in the target groups in a preset pixel image comprises:

[0109] retain the pixel values of all pixel points in the target groups in the preset pixel image, and set the pixel values of pixel points other than the target groups to be empty.

[0110] Figure 7 A flowchart of the region of interest drawing provided by the embodiments of the present application. Combined with the above description of the embodiments of the present application, Figure 7As shown, the preset pixel image is an image corresponding to the full die image, and the number of pixel points and the positions of the pixel points on the preset pixel image correspond to the full die image. Further, if the target group corresponds to multiple groups, the data corresponding to each group needs to be processed separately, and the processing order can be in the order of the number of corresponding pixel points from the most to the least. Specifically, first, the corresponding pixel points of the group with the most number of pixel points are labeled on the preset pixel image (i.e., the pixel values of the pixel points in the current group are retained, and the pixel values of other positions are empty, such as shown by a in FIG. 8). Figure 7 Then, the windowing operation is used to remove isolated or outlier pixel points, for example, "if only the center pixel point has a pixel value other than 0 in a 3x3 window, the pixel point is removed" or "if the outermost pixel points have a pixel value of 0 in a 5x5 window, and only one pixel point has a pixel value other than 0 in a 3x3 window inside the 5x5 window, the pixel point is removed". After removing the isolated pixel points, the remaining pixel points are shown in b in FIG. 8. Figure 7 Further, a circumscribed polygon is constructed based on the remaining pixel points to eliminate diagonal edges, single-pixel jumps, and the like, and the construction result is shown in c in FIG. 8. Figure 7 Finally, in order to cooperate with the detection machine, all the circumscribed polygons in the preset pixel image are split into rectangles, and each rectangle can represent a region of interest, as shown in d in FIG. 8. Figure 7

[0111] In summary, the present application first acquires a full die image of a die to be drawn based on a preset optical mode. Then, based on a preset gray scale range, the target gray scale corresponding to each pixel point in the full die image is determined, and each pixel point is grouped according to the target gray scale, and a first group set is obtained. Finally, the target group is determined from the first group set based on a preset rule, and the region of interest is drawn based on the pixel points corresponding to the target group. In this way, the present application does not rely on wafer design layout, but only distinguishes the types of regions of interest by gray scale difference, thereby realizing automatic drawing of regions of interest and improving the flexibility of region of interest drawing.

[0112] Figure 8 FIG. 8 is a structural schematic diagram of an automatic region of interest drawing device provided by an embodiment of the present application. As shown in FIG. 8, the automatic region of interest drawing device 800 can include: Figure 8

[0113] An acquisition module 801 is configured to acquire a full die image of a die to be drawn based on a preset optical mode.

[0114] A determination module 802 is configured to determine a target gray scale corresponding to each pixel point in the full die image based on a preset gray scale range.

[0115] ​​The grouping module 803 is configured to group each pixel point according to the target gray scale, and obtain a first group set.

[0116] The drawing module 804 is configured to determine a target group from the first group set based on a preset rule, and draw a region of interest based on the pixel points corresponding to the target group.

[0117] In the first group set, the pixel points corresponding to each group correspond to the same target gray scale under each optical mode.

[0118] As an implementation form, to obtain the full-die image of the die to be drawn based on the preset optical mode, the obtaining module 801 specifically includes an obtaining sub-module and a collecting module.

[0119] The obtaining sub-module is configured to obtain a surface image of a wafer to be measured on which the die to be drawn is located based on a preset optical mode.

[0120] The collecting module is configured to collect a full-die image corresponding to the die to be drawn from the surface image based on a line scanning technology.

[0121] As an implementation form, to select the preset optical mode, the device 800 for automatically drawing a region of interest further includes a selection module.

[0122] The selection module is configured to select N optical modes from a set of optical modes as the preset optical modes; the N is an integer greater than or equal to 1.

[0123] Any two optical modes in the set of optical modes are different in at least one of a wavelength band, a polarization, and a focusing mode.

[0124] As an implementation form, to collect the full-die image corresponding to the die to be drawn from the surface image, the collecting module is specifically configured to:

[0125] select K target dies from the surface image; the K is an integer greater than or equal to 1; the target dies and the die to be drawn are the same kind of dies;

[0126] collect full-die images corresponding to the target dies from the surface image based on a line scanning technology.

[0127] In the step of selecting K target dies from the surface image, the step includes:

[0128] selecting K target dies from a middle region of the surface image;

[0129] When the K is greater than 1, the selected target dies are adjacent.

[0130] As an implementation form, the device 800 for automatically drawing the region of interest further comprises an image processing module, for how to perform image processing.

[0131] The image processing module is configured to, when the K is greater than or equal to 2, perform sub-pixel alignment processing on the full-grain image corresponding to each target grain, so that the number of pixels in each full-grain image is the same and completely aligned.

[0132] Merge all full-grain images collected under the same optical mode; the gray value of each pixel point in the merged full-grain image is the median of the gray values of the corresponding pixel points in the full-grain images before merging.

[0133] Perform image noise reduction processing on the merged full-grain image.

[0134] As an implementation form, the drawing module 804 is specifically configured to, for how to determine the target group:

[0135] Determine the number of pixel points corresponding to each group in the first group set;

[0136] Sort the groups according to the number of pixel points from more to less;

[0137] The groups ranked in the first X are taken as the target group; the X is an integer greater than or equal to 1.

[0138] As an implementation form, the drawing module 804 is specifically configured to, for how to draw the region of interest based on the pixel points corresponding to the target group:

[0139] Label all pixel points in the target group in a preset pixel image;

[0140] Eliminate isolated pixel points from the all pixel points, and construct an enclosing polygon based on the remaining pixel points;

[0141] Split the enclosing polygon into rectangles, and draw the region of interest based on the rectangles.

[0142] The labeling of all pixel points in the target group in the preset pixel image comprises:

[0143] Reserve the pixel values of all pixel points in the target group in the preset pixel image, and set the pixel values of pixel points at positions other than the target group to be empty.

[0144] To sum up, the application first acquires a full-die image of a die to be drawn based on a preset optical mode. Then, a target gray scale corresponding to each pixel point in the full-die image is determined based on a preset gray scale range, each pixel point is grouped according to the target gray scale, and a first group set is obtained. Finally, a target group is determined from the first group set based on a preset rule, and a region of interest is drawn based on the pixel points corresponding to the target group. In this way, the application does not rely on a wafer design layout, but only distinguishes the types of regions of interest through gray scale differences, thereby realizing automatic drawing of regions of interest and improving the flexibility of region of interest drawing.

[0145] In addition, the application further provides a device for automatically drawing a region of interest, which comprises a memory for storing a computer program and a processor for implementing the steps of the method for automatically drawing a region of interest as described above when the computer program is executed.

[0146] In addition, the application further provides a readable storage medium having a computer program stored thereon, and the computer program implements the steps of the method for automatically drawing a region of interest as described above when executed by a processor.

[0147] The above description of disclosed embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be apparent to those of ordinary skill in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Accordingly, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of automatically mapping a region of interest, characterized by, The method comprises: acquiring a full-die image of a die to be drawn based on a preset optical mode; determining a target gray scale corresponding to each pixel point in the full-die image based on a preset gray scale range; the preset gray scale range corresponds to X different gray value ranges, each gray value range corresponds to a gray scale; the target gray scale is one of all gray scales; X is a positive integer greater than or equal to 2; grouping each pixel point according to the target gray scale, and obtaining a first group set; the pixel points corresponding to each group in the first group set correspond to only the same target gray scale under each optical mode; determining a target group from the first group set based on a preset rule, and performing region of interest drawing based on the pixel points corresponding to the target group.

2. The method of claim 1, wherein, The method comprises: acquiring a surface image of a wafer to be measured in which the die to be drawn is located based on a preset optical mode; collecting a full-die image corresponding to the die to be drawn from the surface image based on line scanning technology.

3. The method of claim 1, wherein, The method further comprises: selecting N optical modes from an optical mode set as preset optical modes; N is an integer greater than or equal to 1; there is at least one difference between any two optical modes in the optical mode set in terms of waveband, polarization, and focusing mode.

4. The method of claim 2, wherein, The method comprises: selecting K target dies from the surface image; K is an integer greater than or equal to 1; the target dies and the die to be drawn are the same kind of dies; collecting full-die images corresponding to the target dies from the surface image based on line scanning technology.

5. The method of claim 4, wherein, The method comprises: selecting K target dies from the middle region of the surface image; when K is greater than 1, each selected target die is adjacent.

6. The method of claim 4, wherein, The method further comprises: when K is greater than or equal to 2, performing sub-pixel alignment processing on the full-die images corresponding to each target die, so that the number of pixels in each full-die image is the same and completely aligned; merging all full-die images collected under the same optical mode; the gray values of each pixel point in the merged full-die image are the median values of the gray values of the corresponding pixel points in the full-die images before merging; performing image noise reduction processing on the merged full-die image.

7. The method of claim 1, wherein, The pixel points corresponding to each group in the first group set correspond to only the same target gray scale under each optical mode.

8. The method of claim 1, wherein, The method comprises: determining the number of pixel points corresponding to each group in the first group set; sorting the groups according to the number of pixel points from more to less; selecting the first X groups as target groups; X is an integer greater than or equal to 1.

9. The method of claim 1, wherein, The method comprises: labeling all pixel points in the target groups in a preset pixel image; the number of pixel points and the positions of each pixel point in the preset pixel image correspond to the full-die image; Eliminate isolated pixel points from the total pixel points, and construct an external polygon based on the remaining pixel points; Split the external polygon into rectangles, and perform region of interest drawing based on the rectangles.

10. The method of claim 9, wherein, The step of labeling all pixel points in the target group in the preset pixel image comprises: Reserve the pixel values of all pixel points in the target group in the preset pixel image, and set the pixel values of pixel points outside the target group to be empty.

11. An apparatus for automatically delineating a region of interest, the apparatus comprising: Comprise: An acquisition module is configured to acquire a full die image of a die to be drawn based on a preset optical mode; A determination module is configured to determine a target gray scale corresponding to each pixel point in the full die image based on a preset gray scale range; The preset gray scale range corresponds to X different gray value ranges, and each gray value range corresponds to a gray scale; the target gray scale is one of all gray scales; X is a positive integer greater than or equal to 2; A grouping module is configured to group each pixel point according to the target gray scale and obtain a first group set; The pixel points corresponding to each group in the first group set correspond to only the same target gray scale under each optical mode; A drawing module is configured to determine a target group from the first group set based on a preset rule, and perform region of interest drawing based on the pixel points corresponding to the target group.

12. An apparatus for automatically mapping a region of interest, the apparatus comprising: Comprise: A memory is configured to store a computer program; A processor is configured to implement the steps of the method for automatically drawing a region of interest according to any one of claims 1 to 10 when executing the computer program.

13. A readable storage medium, characterized by, The computer program is stored on the readable storage medium, and the computer program is executed by the processor to implement the steps of the method for automatically drawing a region of interest according to any one of claims 1 to 10.

Images