TW202219900A - Image analysis method and image analysis system - Google Patents

Abstract

The image analysis method of the present invention includes following steps: obtaining an image of a multi-layer structure provided by an electronic microscope and displaying the image of the multi-layer structure through a display device, wherein the image of the multi-layer structure is a grey-scale image; setting a measurement line segment, wherein the measurement line segment extends along a first direction; detecting a grey-scale distribution within the measurement line segment corresponding to the image of the multi-layer structure along the measurement line segment; analyzing the grey-scale distribution to determine a plurality of dark layer thicknesses and a plurality of light layer thickness according to a threshold range.

Description

Image analysis method and image analysis system

The present invention relates to an analysis method, and more particularly, to an image analysis method and an image analysis system.

In the semiconductor process, the size of the device will affect the change of electrical properties. Therefore, in terms of size requirements, it is often necessary to be very precise. When measuring semiconductor elements, electron microscopes with magnifying inspection functions, such as transmission electron microscopes and scanning electron microscopes, are usually used. However, when using an electron microscope to measure the size of an image of a component, the edge points of each region are often manually set one by one to obtain the size of each region, which takes a long time. In view of this, solutions of several embodiments will be proposed below.

The invention provides an image analysis method and an image analysis system, which can automatically measure the thickness of each layer of a multi-layer structure image according to a set measurement line segment.

The image analysis method of the present invention includes the following steps: obtaining a multi-layer structure image provided by an electron microscope, and displaying the multi-layer structure image through a display device, wherein the multi-layer structure image is a gray-scale image; setting a measurement line segment on the multi-layer structure image, wherein the The measurement line segment extends toward the first direction; the gray level distribution of the multilayer structure image corresponding to the measurement line segment is detected along the measurement line segment; and the gray level distribution is analyzed to determine a plurality of depths in the multilayer structure image according to the threshold range Color layer thickness and multiple light color layer thicknesses.

The image analysis system of the present invention includes an electron microscope, a display device, and an image analysis device. Electron microscopy is used to provide images of multilayer structures. The display device is used for displaying the multi-layer structure image. The image analysis device is coupled to the electron microscope and the display device, so as to obtain the multilayer structure image provided by the electron microscope and output the multilayer structure image to the display device. The image analysis device includes a storage device and a processor. The storage device includes an image analysis module. The processor is coupled to the storage device. The processor inputs the multilayer structure image to the image analysis module. The processor sets the measurement line segment on the multilayer structure image, wherein the measurement line segment extends toward the first direction. The processor detects, through the image analysis module, the grayscale distribution of the multi-layer structure image corresponding to the measurement line segment along the measurement line segment. The processor analyzes the grayscale distribution through the image analysis module, so as to determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image according to the threshold range.

Based on the above, the image analysis method and image analysis system of the present invention can automatically measure the thickness of each layer of the multi-layer structure image according to the set measurement line segment, thereby saving a lot of time required for manual operations.

In order to make the above-mentioned features and advantages of the present invention more obvious and easy to understand, the following embodiments are given and described in detail with the accompanying drawings as follows.

In order to make the content of the present invention more comprehensible, the following specific embodiments are given as examples according to which the present invention can indeed be implemented. In addition, where possible, elements/components/steps using the same reference numerals in the drawings and embodiments represent the same or similar parts.

Also, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be construed as having meanings consistent with their meanings in the context of the related art and the present invention, and are not to be construed as idealized or excessive Formal meaning, unless expressly defined as such herein.

FIG. 1 is a schematic diagram of an image analysis system according to an embodiment of the present invention. Referring to FIG. 1 , the image analysis system 100 may include an image analysis device 101 , an electron microscope 140 and a display device 150 . The electron microscope 140 can be used to provide images of multilayer structures by photographing semiconductor process objects (semiconductor products). The multilayer structure image is an electron microscope image, and is a grayscale image. The multi-layer structure image may include multi-layer semiconductor structure layers of different materials, and the gray scale distribution of the image of the multi-layer semiconductor structure layers may be determined according to different semiconductor materials.

In this embodiment, the display device 150 can be used to display a multi-layer structure image. The image analysis device 101 can be coupled to the electron microscope 140 and the display device 150 to obtain the multilayer structure image provided by the electron microscope 140 and output the multilayer structure image to the display device 150 . The image analysis device 101 may include a processor 110 and a storage device 120 . The storage device 120 may include an image analysis module 121 . The processor 110 may be coupled to the storage device 120 . In this embodiment, the image analysis device 101 may be an independent computer device or a cloud server, but the invention is not limited thereto.

In this embodiment, the processor 110 can input the multi-layer structure image to the image analysis module 121, and the processor 110 can set a measurement line segment on the multi-layer structure image, wherein the measurement line segment extends toward the stacking direction of the multi-layer structure. In this embodiment, the manner in which the processor 110 sets the measurement line segment may include manual setting or automatic setting. In this regard, manual setting may, for example, refer to setting instructions or parameters provided by the input device of the image analysis system 100 (eg, an input operation performed by a user) to set the position of the measurement line segment, but the invention is not limited thereto. The automatic setting may, for example, refer to the automatic setting by the image analysis system 100 according to the boundary range of the picture or the automatic setting according to a preset condition, but the present invention is not limited to this.

Next, the processor 110 can detect the gray-scale distribution of the multilayer structure image corresponding to the measurement line segment along the measurement line segment through the image analysis module 121 . In addition, the processor 110 can analyze the grayscale distribution through the image analysis module 121 to determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image according to the threshold. In this way, the image analysis system 100 can automatically measure the thickness of each layer of the multi-layer structure image according to the set measurement line segment, thereby saving a lot of time required for manual operations.

In this embodiment, the processor 110 may include, for example, a central processing unit (Central Processing Unit, CPU), a microprocessor (Microprocessor Control Unit, MCU), or a Field Programmable Gate Array (Field Programmable Gate Array, FPGA). Inventions are not limited to this.

In this embodiment, the storage device 120 may include, for example, random-access memory (RAM), read-only memory (ROM), optical disc (optical disc), magnetic disk (Magnetic disk) ), Hard drive, Solid-state drive, Flash drive, Security digital (SD) card, Memory stick, Compact Flash ( Compact flash; CF) card or any type of storage device, but the present invention is not limited to this. The storage device 120 can store the image analysis module 121 and the related image data, the related analysis results and data, and the display interface described in the various embodiments for the processor 110 to read and execute them.

In this embodiment, the electron microscope 140 may include, for example, a scanning electron microscope (Scanning Electron Microscope; SEM) and a transmission electron microscope (Transmission electron microscope; TEM), but the invention is not limited thereto. In this embodiment, the display device 150 may include, for example, various electronic devices having a display function. In addition, in another embodiment, the display device 150 can be disposed in the image analysis device 101 , so that the image analysis device 101 can be, for example, a computer device with a display function.

FIG. 2 is a flowchart of an image analysis method according to an embodiment of the present invention. FIG. 3 is a schematic diagram of an image of a multi-layer structure according to an embodiment of the present invention. FIG. 4 is a schematic diagram of gray scale distribution according to an embodiment of the present invention. Referring to FIGS. 1 to 4 , in this embodiment, the image analysis system 100 may perform steps S210 to S250 of the following image analysis method. In step S210 , the image analysis device 101 can obtain the multilayer structure image 300 provided by the electron microscope 140 . In step S220 , the image analysis device 101 can display the multi-layer structure image 300 through the display device 150 . In step S230 , the image analysis module 121 may set the measurement line segment 330 on the multilayer structure image 300 . In step S240 , the image analysis module 121 detects the grayscale distribution 400 of the multilayer structure image 300 corresponding to the measurement line segment 330 along the measurement line segment 330 . In step S250 , the image analysis module 121 analyzes the grayscale distribution 400 to determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image 300 according to the threshold ranges 461 and 462 .

Specifically, the multilayer structure image 300 can be a grayscale image, and the first direction P1 is perpendicular to the second direction P2. In addition, the multi-layer structure image 300 may include a plurality of dark layer images 310-1 to 310-4 and a plurality of light layer images 320-1 to 320-5, the plurality of dark layer images 310-1 ~310-4 and the plurality of light layer images 320-1 ~ 320-5 are staggered along the first direction P1, the plurality of dark layer images 310-1 ~ 310-4 and the The plurality of light-color layer images 320-1 to 320-5 extend along the second direction P2, respectively. In this embodiment, the plurality of dark layer images 310-1 to 310-4 may be the first type of semiconductor material layers, and the plurality of light layer images 320-1 to 320-5 may be the first type of semiconductor material layers. Two types of semiconductor material layers, wherein the first type of semiconductor material layer is different from the second type of semiconductor material layer. In this embodiment, the plurality of dark layer images 310-1 to 310-4 and the plurality of light layer images 320-1 to 320-5 may further include a plurality of white thin layer images, respectively. 340-1 to 340-7, and the plurality of white thin layer images 340-1 to 340-7 may be a third type of semiconductor material layers different from the first type and the second type.

1 to 4 , in step S230 , the processor 110 can execute the image analysis module 121 to set the measurement line segment 330 on the multilayer structure image 300 (as described above manually or automatically). In step S240 , the processor 110 may detect a plurality of consecutive pixels of the multilayer structure image 300 corresponding to the measurement line 330 along the measurement line 330 to obtain a plurality of grayscale values of the plurality of pixels. Furthermore, the processor 110 can establish the gray-scale distribution 400 shown in FIG. 4 according to the plurality of gray-scale values. In step S250 , the processor 110 may analyze the grayscale distribution 400 and determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image 300 according to the preset threshold ranges 461 and 462 . In this regard, as shown in FIG. 4 , the value of the horizontal axis of the grayscale distribution 400 may correspond to each pixel along the first direction P1 on the measurement line segment 330 , and the value of the vertical axis may represent that the multi-layer structure image 300 corresponds to the measurement Grayscale value of each pixel on line segment 330 .

In this embodiment, the threshold range 461 (also called the first threshold range) can be set to, for example, 0-15, and the threshold range 462 (also called the second threshold range) can be set to, for example, 45-90, but the present invention Not limited to this. That is to say, when the grayscale value of a certain pixel is between 0 and 15, the certain pixel can be regarded as a dark pixel. Conversely, when the grayscale value of another pixel is between 45 and 90, the other pixel can be regarded as a light-colored pixel. For example, in step S250, the processor 110 may determine pixels whose grayscale values are between the threshold range 461 as dark pixels. Also, the processor 110 may determine pixels whose grayscale values are between the threshold range 462 as light-colored pixels. Next, the processor 110 may count the number of dark pixels corresponding to the thickness of each dark layer or the thickness of each light layer or the number of light pixels on the measurement line segment 330 to obtain the thickness or thickness of each dark layer respectively. layer thickness. That is, each dark layer thickness corresponds to the number of dark pixels of each dark layer thickness, and each light layer thickness corresponds to the number of light pixels of each light layer thickness.

In this embodiment, when the processor 110 calculates the number of dark pixels in each dark layer and the number of light pixels in each light layer, the processor 110 can instantly convert them into corresponding thickness parameters, to output the measurement results. For example, one pixel may correspond to 1 nanometer. In one embodiment, assuming that the 6th to 44th pixels are light-colored pixels, the processor 110 can obtain a light-colored layer image corresponding to a certain layer with a thickness of 39 nm (44−6+1=39). . However, the corresponding relationship between the pixels and the length can be adjusted by manual setting or automatic detection, and the present invention is not limited thereto.

In addition, in this embodiment, the overall measurement result can be presented in the form of the following table (1) and displayed on the display device 150, but the present invention is not limited to this. In this embodiment, the layer number may represent a plurality of dark layer images 310 - 1 to 310 - 4 or a plurality of light layers in which the measurement line segment 330 and the multilayer structure image 300 intersect in sequence along the first direction P1 Numbering of images 320-1 to 320-5. In this embodiment, the average grayscale value may represent the dark layer images 310-1 to 310-4 or the light layer image 320 where the measurement line segment 330 and the multilayer structure image 300 intersect in sequence along the first direction P1. -1~320-5 The average value of the grayscale values of the corresponding multiple pixels. In this embodiment, the measured thickness may represent the thickness of the dark layer or the thickness of the light layer corresponding to the layer number. layer number Average grayscale value Measuring thickness 1 8 38 2 80 39 3 5 38 4 83 35 Table 1)

It should be noted that, in one embodiment, the multi-layer structure image 300 may include a plurality of dark layer images 310-1 to 310-4 and light layer images 320-1 to 320-5 with different grayscale value ranges. , and the processor 110 can correspondingly set threshold ranges 461 , 462 or other threshold ranges to measure the thickness. Moreover, in one embodiment, in the multi-layer structure image 300, the dark layer images 310-1 to 310-4 and the light layer images 320-1 to 320-5 of a plurality of different grayscale value ranges are along the The first direction P1 may be randomly arranged, but not limited to the dark layer images 310-1 to 310-4 and the light layer images 320-1 to 320-5 being staggered along the first direction P1.

FIG. 5 is a schematic diagram of a display interface according to an embodiment of the present invention. Referring to FIG. 3 and FIG. 5 , the display interface 500 may include a multi-layer structure image 510 , a toolbar 520 and a measurement result 530 . For the description of the multilayer structure image 510 , reference may be made to the multilayer structure image 300 in FIG. 3 , and details are not repeated here. In this embodiment, the tool bar 520 may include buttons such as a magnifying glass, moving a display range, and setting a measurement line segment, etc., so as to provide a user to operate, but the invention is not limited thereto. For example, the user can use the button on the toolbar 520 to set the measurement line segment to set the desired measurement line segment 330 by himself. Moreover, the measurement result 530 may include, for example, the number of the layer, and the thickness of the dark layer corresponding to the dark layer or the thickness of the light layer corresponding to the light layer. It is worth noting that, in the measurement result 530, the thickness of the dark layer and the thickness of the light layer each correspond to a same layer number. In another embodiment, the measurement result 530 may be in the format of Table (1), and the layer number corresponds to only one dark layer thickness or light layer thickness. In this embodiment, the display interface 500 can be displayed on the display device 150 so that the user can view the measurement result 530 in real time.

FIG. 6 is a flowchart of an image analysis method according to an embodiment of the present invention. Referring to FIGS. 1 to 4 and FIG. 6 , step S250 in FIG. 2 can be implemented, for example, in the manner of FIG. 6 . In this embodiment, when the processor 110 executes the image analysis module 121 to analyze the grayscale distribution 400, it can be divided into the following four steps: step S610 (finding the thickness of the dark layer or the starting point of the light layer), step S610 S620 (find the end point of the thickness of the dark layer or the thickness of the light layer), step S630 (check the thickness range), and step S640 (check whether the thicknesses overlap).

It is worth noting that the grayscale values of the white thin layer images 340-1 to 340-7 are higher than those of the dark layer images 310-1 to 310-4 and the light layer images 320-1 to 320-5. grayscale value. Therefore, when the processor 110 analyzes the grayscale distribution 400, whether it is from the white thin layer images 340-1 to 340-7 into the dark layer images 310-1 to 310-4 or the light layer image 320-1 ~320-5, the grayscale value will decrease. That is, the slope value of two consecutive grayscale values will be negative. Next, the processor 110 only needs to determine whether the interval formed by the two consecutive grayscale values includes the upper limit of the threshold ranges 461 and 462, and whether the second consecutive grayscale value belongs to the dark layer images 310-1 to 310 The threshold value range 461 of -4 or the threshold value range 462 of the light layer images 320-1 to 320-5 can indicate the starting point of the thickness of the dark layer or the thickness of the light layer. That is, the processor 110 determines that the interval formed by the two consecutive grayscale values includes the upper limit value (also referred to as the first upper limit value) of the threshold range 461, and determines that the second pixel of the two consecutive pixels is In the case of dark pixels, the second pixel of the corresponding two consecutive pixels is marked as the starting point of the dark layer corresponding to the thickness of the dark layer. Similarly, the processor 110 determines that the interval formed by the two consecutive grayscale values includes the upper limit value (also referred to as the second upper limit value) of the threshold range 462, and determines that the second pixel of the two consecutive pixels is shallow When color pixels are selected, the second pixel of the corresponding two consecutive pixels is marked as the starting point of the light-colored layer corresponding to the thickness of the light-colored layer.

Moreover, when the processor 110 analyzes the gray-scale distribution 400, whether it is from the dark layer images 310-1 to 310-4 or the light layer images 320-1 to 320-5 to enter the white thin layer image 340-1 ~340-7, the grayscale value will increase. That is, the slope value of two consecutive grayscale values will be positive. Next, the processor 110 only needs to determine whether the interval formed by the two consecutive grayscale values includes the upper limit of the threshold ranges 461 and 462, and whether the second consecutive grayscale value does not belong to the dark layer image 310-1~ The threshold value range 461 of 310-4 or the threshold value range 462 of the light-colored layer images 320-1 to 320-5 can indicate the end point of the thickness of the dark-colored layer or the thickness of the light-colored layer. That is to say, when the processor 110 determines that the interval formed by the two consecutive grayscale values includes the first upper limit value of the threshold range 461, and determines that the first pixel of the two consecutive pixels is a dark pixel, the marker corresponding to The first pixel of two consecutive pixels is the end point of the dark layer corresponding to the thickness of the dark layer. Similarly, when the processor 110 determines that the interval formed by two consecutive grayscale values includes the second upper limit value of the threshold range 462, and determines that the first pixel of the two consecutive pixels is a light-colored pixel, the processor 110 marks the The first pixel of the corresponding two consecutive pixels is the end point of the light-colored layer corresponding to the thickness of the light-colored layer.

In this embodiment, the processor 110 may set a falling threshold value, and determine whether the slope value is smaller than the falling threshold value, so as to mark the starting point of the thickness of the dark layer or the thickness of the light layer. In this embodiment, the processor 110 may set a rising threshold value, and determine whether the slope value is greater than the rising threshold value, so as to mark the end point of the thickness of the dark layer or the thickness of the light layer. For example, both the falling threshold and the rising threshold may be set to zero, but the invention is not limited thereto. That is, the processor 110 can determine whether the slope value is a negative value (less than zero), so as to indicate the starting point of the thickness of the dark layer or the thickness of the light layer. Furthermore, the processor 110 can determine whether the slope value is a positive value (greater than zero) to mark the end point of the thickness of the dark layer or the thickness of the light layer. In one embodiment, the falling threshold value and the rising threshold value can be respectively set to the same or different values according to design requirements, but the invention is not limited thereto.

For example, in step S610, when analyzing the grayscale distribution 400, the processor 110 may calculate the slope value of two consecutive grayscale values, and determine the difference between the two consecutive grayscale values according to the change of the slope value and the difference between the two consecutive grayscale values. Whether the second grayscale value falls within the threshold range 461 corresponding to the dark layer images 310-1~310-4 or the threshold range 462 corresponding to the light layer images 320-1~320-5, so that the two consecutive The pixel corresponding to the second grayscale value of the first grayscale value is marked as the starting point of the dark layer. That is to say, the processor 110 determines whether the interval formed by the two consecutive grayscale values includes the upper limit value of the threshold range 461 and 462, so as to mark the pixel corresponding to the second grayscale value of the two consecutive grayscale values The starting point for the dark or light layer.

In step S620, when analyzing the grayscale distribution, the processor 110 may calculate the slope value of two consecutive grayscale values, and determine the second grayscale of the two consecutive grayscale values according to the change of the slope value and the Whether the value leaves the threshold value range 461 corresponding to the dark layer images 310-1~310-4 or the threshold value range 462 corresponding to the light layer images 320-1~320-5, so that the first two consecutive grayscale values A pixel corresponding to a grayscale value is marked as the end point of the dark layer or the light layer. That is, the processor 110 determines whether the interval formed by the two consecutive grayscale values includes the upper limit value of the threshold range 461 and 462, so as to mark the pixel corresponding to the first grayscale value of the two consecutive grayscale values It is the end point of the dark layer or the light layer.

Next, in step S630, the processor 110 calculates the number of pixels between the start point and the end point corresponding to each dark layer thickness or light color layer thickness. In this way, the processor 110 can calculate the corresponding dark layer thickness or light layer thickness according to the number of pixels between the starting point and the end point corresponding to each dark layer thickness or light layer thickness.

Also, the processor 110 may check the pixels to discard incorrect starting points. For example, the processor 110 may check whether the average value of the grayscale values of all pixels between the starting point and the ending point of each dark layer thickness is within the threshold range 461 corresponding to the dark layer images 310-1 to 310-4 or It is checked that the average value of the grayscale values of all pixels between the starting point and the ending point corresponding to the thickness of the light-colored layer is within the threshold range 462 corresponding to the light-colored layer images 320-1 to 320-5. After the checking result is correct, the processor 110 calculates the number of pixels between the start point and the end point corresponding to each dark layer thickness or light color layer thickness. On the contrary, after the checking result is incorrect, the processor 110 may discard the current starting point, and re-determine a starting point as a new starting point.

Finally, in step S640, the processor 110 may check whether adjacent dark layer thicknesses or light color layer thicknesses correspond to the same pixel to check whether the thicknesses overlap, and regard the overlapping thicknesses as the same dark layer thickness or light layer thickness. That is, after the processor 110 checks that the thicknesses of the adjacent dark layers or the thicknesses of the light layers correspond to the same pixel, the processor 110 can combine the thicknesses of the adjacent light layers or the thicknesses of the dark layers into a single thickness of the light layers or dark layer thickness. In this way, the processor 110 can correct the overlapping of the dark layer or the light layer caused by the interference of the noise, so as to obtain the correct thickness of the dark layer or the light layer. It is worth noting that, in an embodiment, step S640 may be omitted, and the thickness calculated in step S630 is directly used as the thickness of the dark layer or the thickness of the light layer.

FIG. 7 is a schematic diagram of gray scale distribution of a multi-layer structure image according to an embodiment of the present invention. Referring to FIG. 1 and FIG. 7 , in this embodiment, the grayscale distribution 700 of the multi-layer structure image may only include a plurality of dark layer images 710-1~710-4 and a plurality of light layer images 720-1~ 720-3, and may not include white thin layer images 340-1 to 340-7 as shown in FIG. 3 . Similar to the foregoing method, when determining the starting point or end point of the thickness of the dark layer or the thickness of the light layer, the processor 110 can respectively determine whether the second consecutive grayscale values belong to the dark layer images 710 - 1 to 710 , respectively. The threshold value range 761 of -4 or the threshold value range 762 of the light layer images 720-1 to 720-3. In other words, the processor 110 can determine whether the interval formed by the two consecutive grayscale values includes the upper limit of the threshold range 761, so as to mark the start or end point of the thickness of the dark layer. Moreover, the processor 110 can determine whether the interval formed by the two consecutive grayscale values includes the upper limit value of the threshold range 762, so as to mark the starting point or the ending point of the thickness of the light color layer. Then, the processor 110 may enter or leave the threshold range 761 of the dark layer images 710-1 to 710-4 or the light layer images 720-1 to 720-3 according to whether the second pixel of the two consecutive pixels The threshold value range 762 is used to determine whether the starting point or the ending point belongs to the thickness of the dark layer or the thickness of the light layer. In addition, the detailed calculation method of the thickness is the same as that described above, and will not be repeated here.

In addition, the processor 110 may determine the starting point or the ending point of the thickness of the dark layer or the thickness of the light layer according to the positive or negative of the slope value, but the present invention is not limited thereto. Specifically, when going from a dark layer to a light layer, the grayscale value increases, and when going from a light layer to a dark layer, the grayscale value decreases. That is, the processor 110 may determine that the second pixel of the two consecutive pixels belongs to a dark pixel or a light pixel according to the slope value of the two consecutive grayscale values. For example, the processor 110 may determine the starting point 710a of the dark layer thickness according to the negative slope value, and the processor 110 may determine the end point 710b of the dark layer thickness according to the positive slope value. Moreover, the processor 110 can determine the starting point 720a of the thickness of the light color layer according to the positive slope value, and the processor 110 can determine the end point 720b of the thickness of the dark layer according to the negative slope value.

To sum up, the image analysis method and image analysis system of the present invention can automatically and quickly measure the thickness of each layer of the multi-layer structure image according to the set measurement line segment, thereby saving the need for manual thickness measurement layer by layer. a lot of time. In addition, the image analysis method and the image analysis system of the present invention can effectively avoid the influence of image noise or material impurities, and can accurately measure the thickness of each layer in the multi-layer structure image.

Although the present invention has been disclosed above by the embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, The protection scope of the present invention shall be determined by the scope of the appended patent application.

100: Image Analysis System 101: Image analysis device 110: Processor 120: Storage Device 121: Image Analysis Module 140: Electron Microscopy 150: Display device S210, S220, S230, S240, S250, S610, S620, S630, S640, S650: Steps 300: Multilayer Structure Image 310-1~310-4, 710-1~710-4: Dark layer images 320-1~320-5, 720-1~720-3: Light layer image 330: Measuring line segment 340-1~340-7: White thin layer image 400: Grayscale distribution 461, 462, 761, 762: Threshold range 500: Display interface 510: Multilayer Structure Image 520: Toolbar 530: Measurement results 700: Grayscale distribution of multi-layer structure images 710a, 720a: starting point 710b, 720b: End point P1: first direction P2: Second direction

FIG. 1 is a schematic diagram of an image analysis system according to an embodiment of the present invention. FIG. 2 is a flowchart of an image analysis method according to an embodiment of the present invention. FIG. 3 is a schematic diagram of an image of a multi-layer structure according to an embodiment of the present invention. FIG. 4 is a schematic diagram of gray scale distribution according to an embodiment of the present invention. FIG. 5 is a schematic diagram of a display interface according to an embodiment of the present invention. FIG. 6 is a flowchart of an image analysis method according to an embodiment of the present invention. FIG. 7 is a schematic diagram of gray scale distribution of a multi-layer structure image according to an embodiment of the present invention.

100: Image Analysis System

101: Image analysis device

110: Processor

120: Storage Device

121: Image Analysis Module

140: Electron Microscopy

150: Display device

Claims (10)

An image analysis method comprising: obtaining a multi-layer structure image provided by an electron microscope, and displaying the multi-layer structure image through a display device, wherein the multi-layer structure image is a grayscale image; setting a measurement line segment on the multilayer structure image, wherein the measurement line segment extends toward a first direction; detecting a grayscale distribution of the multilayer structure image corresponding to the measurement line segment along the measurement line segment; and The grayscale distribution is analyzed to determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image according to a threshold range. The image analysis method according to claim 1, wherein the multi-layer structure image includes a plurality of dark layer images and a plurality of light layer images, and the dark layer images and the light layer images are along the The first direction is staggered, and the dark layer images and the light layer images respectively extend along a second direction, wherein the first direction is perpendicular to the second direction. The image analysis method of claim 1, wherein the grayscale distribution includes a distribution of grayscale values of the multilayer structure image corresponding to a plurality of pixels on the measurement line segment, and analyzing the grayscale distribution includes comparing the grayscale distribution The grayscale values of the pixels and the threshold range are determined to determine a plurality of dark pixels corresponding to the thicknesses of the dark layers and a plurality of light pixels corresponding to the thicknesses of the light layers. The image analysis method according to claim 3, wherein the threshold range includes a first threshold range and a second threshold range, wherein analyzing the grayscale distribution includes: When judging that the grayscale values belong to the first threshold range, determine that the pixels corresponding to the grayscale values belong to the dark pixels; and When determining that the gray-scale values belong to the second threshold range, it is determined that the pixels corresponding to the gray-scale values belong to the light-colored pixels. The image analysis method of claim 4, wherein the thicknesses of the dark layers correspond to the number of the dark pixels of the thicknesses of the dark layers, and the thicknesses of the light layers correspond to the thicknesses of the light layers The number of light-colored pixels. The image analysis method according to claim 4, wherein analyzing the grayscale distribution comprises: When judging that an interval formed by two consecutive gray-scale values includes a first upper limit value of the first threshold range, and judging that the second pixel of the two consecutive pixels is the dark pixel, mark The second one of the corresponding two consecutive pixels is a starting point of a dark layer corresponding to the thickness of the dark layer; and When judging that an interval formed by two consecutive grayscale values includes a second upper limit value of the second threshold range, and judging that the second one of the two consecutive pixels is the light-colored pixel, mark The second pixel of the corresponding two consecutive pixels is the starting point of a light-colored layer corresponding to the thickness of the light-colored layer. The image analysis method according to claim 5, wherein analyzing the grayscale distribution comprises: When judging that an interval formed by two consecutive grayscale values includes the first upper limit value of the first threshold range, and judging that the first one of the two consecutive pixels is the dark pixel, mark The first one of the corresponding two consecutive pixels is a dark layer end point of the corresponding dark layer thickness; and When judging that an interval formed by two consecutive gray-scale values includes the second upper limit value of the second threshold range, and judging that the first one of the two consecutive pixels is the light-colored pixel, mark the The first pixel of the corresponding two consecutive pixels is the end point of a light-colored layer corresponding to the thickness of the light-colored layer. The image analysis method of claim 7, wherein the dark layer thicknesses correspond to the number of dark pixels between the dark layer starting point and the dark layer ending point of the dark layer thicknesses and the light Color Layer Thickness The number of the light-colored pixels between the light-colored layer start point and the light-colored layer end point of the light-colored layer thicknesses. The image analysis method according to claim 8, further comprising: after checking that the thickness of the adjacent dark layer or the thickness of the light layer corresponds to the same pixel, analyzing the thickness of the adjacent light layer or the thickness of the dark layer The layer thicknesses are combined into a single thickness of the light-colored layer or the thickness of the dark-colored layer. An image analysis system, comprising: an electron microscope for providing an image of the multilayer structure; a display device for displaying the multi-layer structure image; and An image analysis device coupled to the electron microscope and the display device to obtain the multilayer structure image provided by the electron microscope and output the multilayer structure image to the display device, the image analysis device comprising: a storage device including an image analysis module; and a processor coupled to the storage device, wherein The processor inputs the multilayer structure image to the image analysis module, The processor sets a measurement line segment on the multilayer structure image, wherein the measurement line segment extends toward the first direction, The processor detects a gray-scale distribution of the multilayer structure image corresponding to the measurement line segment along the measurement line segment through the image analysis module, The processor analyzes the grayscale distribution through the image analysis module to determine the thicknesses of the dark layers and the thicknesses of the light layers in the multilayer structure image according to a threshold range.

Images