Vapor deposition equipment substrate state identification method, vapor deposition equipment substrate state identification system and vapor deposition equipment

Abstract

The invention provides a vapor deposition equipment substrate state identification method and system, and vapor deposition equipment, and the method comprises the steps: shooting an image of a tray when a reaction cavity is at an initial temperature and each node temperature, comparing the image corresponding to each node temperature with the image corresponding to the initial temperature, obtaining noise data corresponding to each node temperature, and carrying out the recognition of the state of the substrate of the vapor deposition equipment; in the process of processing the substrate, in the temperature rising process of the reaction cavity, the images of the substrate and the tray when the temperature rises to any node temperature are obtained, the noise data corresponding to the node temperature are combined for processing, then the denoised image is obtained, and the denoised image is clearer than an original image; therefore, the accuracy of substrate state recognition can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly relates to a method and system for identifying the state of a substrate of a chemical vapor deposition (CVD) apparatus and a CVD apparatus. Background Art

[0002] In the process of semiconductor chip production, a large number of microfabrications are required. Among them, a common method is to use a chemical vapor deposition (CVD) process to process semiconductor wafers based on the principle of its vacuum reaction chamber. During the deposition process of a semiconductor substrate, it is necessary to closely monitor the process to ensure that the deposition process results are well controlled.

[0003] Usually, an image capturing device is used to take pictures of the internal substrate through the observation window of the reaction chamber to identify the state of the substrate and determine whether the process needs to be adjusted. Therefore, obtaining a clear internal image to accurately identify the substrate state is of great significance for improving the quality of thin film deposition.

[0004] However, the temperature in the reaction chamber of the CVD apparatus is as high as over 1000 °C and can even reach 1600 °C. The noise generated in such a high-temperature environment will cause the captured image of the interior of the reaction chamber to be blurred, as Figure 1 shown, making it impossible to accurately identify the state of the substrate. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for identifying the state of a substrate of a CVD apparatus and a CVD apparatus, which perform denoising processing on the blurred image in the high-temperature environment inside the captured reaction chamber to obtain a clear image for identifying the state of the substrate.

[0006] To achieve the above purpose, the present invention is realized through the following technical solutions:

[0007] Record the initial temperature T of the reaction chamber 0 , and obtain an initial image I of at least a part of the upper surface of the tray at the initial temperature T 0 through the observation window on the reaction chamber; 0

[0008] Record n node temperatures T 0 during the process of the reaction chamber rising from the initial temperature T 1 to the final temperature, and obtain a background image I of at least a part of the upper surface of the tray corresponding to each node temperature through the observation window on the reaction chamber n ; n

[0009] The initial image I 0 and each background image I nMultiple frequency data and their corresponding amplitude data are obtained through Fourier transform. For each background image I n and the initial image I 0 at the corresponding frequency, the difference in amplitude is taken to obtain the noise data D n ;

[0010] Place the substrate on the tray, and obtain the original image R including at least a part of the upper surface of the tray and at least a part of the upper surface of the substrate at each of the node temperatures n , and for each of the original images R n after Fourier transform, the amplitude data at the corresponding frequency is subtracted from the noise data D n to obtain the denoised data P n , where n > 0;

[0011] Use the inverse Fourier transform to restore all the denoised data P n to the denoised image, and identify different states of the substrate

[0012] Optionally, the method further includes:

[0013] Record i node temperatures T n+1 ~T n+i from the final temperature to the end of the process of the reaction chamber and the corresponding background images I n+i , and calculate the noise data D n+i , where i > 1;

[0014] Place the substrate on the tray, record the original image R at each node temperature n+i , and calculate the corresponding denoised data P n+i to obtain the denoised image, and identify the substrate state of the substrate during the process

[0015] Optionally, the state includes substrate offset, number of contamination particles, and / or substrate warping

[0016] 4. The method for identifying the substrate state of a chemical vapor deposition device according to claim 3, wherein at any node temperature, the image capturing device and the capturing position are adjusted to capture the original images including different positions of the upper surface of the tray and the upper surface of the substrate

[0017] Optionally, the tray includes a disk pit for accommodating the substrate, and the denoised image includes the edges of the substrate and the disk pit, which are used to identify the offset of the substrate relative to the center of the disk pit

[0018] Optionally, the tray includes a rotatable small tray for accommodating the substrate, and the denoised image includes the edge of the small tray, which is used to identify the offset of the small tray relative to the large tray

[0019] Optionally, the denoised image includes a partial area on the substrate for determining the number of particulate contaminants.

[0020] Optionally, during the process, it also includes adjusting the lens height of the image capturing device to obtain multiple denoised images, and recording the distance of the lens adjustment corresponding to the image with the highest clarity for determining the amount of substrate warpage.

[0021] Optionally, the method further includes:

[0022] After adjusting the gas blowing or temperature of the reaction chamber according to the recognition results of substrate offset, the number of contaminant particles, and / or the substrate warpage state, the original image R including at least a partial upper surface of the tray and at least a partial upper surface of the substrate at the node temperature is executed again n , for each of the original images R n The amplitude data at the corresponding frequency after Fourier transform and the noise data D n Are subtracted to obtain the denoised data P n , and the denoised data P is restored to a denoised image by using inverse Fourier transform n To identify different states of the substrate.

[0023] Optionally, the initial image I 0 Includes an initial visible light image I 01 And an initial infrared image I 02 , the background image I n Includes a background visible light image I n1 And a background infrared image I n2 , the noise data D n Includes visible light image noise data D n1 And infrared image noise data D n2 ;

[0024] The original image R n Includes an original visible light image R n1 And an original infrared image R n2 , the denoised data P n Includes visible light denoised data P n1 And infrared denoised data P n2 , and the visible light denoised data P is restored to a visible light image and an infrared image by using inverse Fourier transform n1 And the infrared denoised data P n2 Are restored to a visible light image and an infrared image, and the visible light image and the infrared image are fused to obtain a denoised image.

[0025] Optionally, after restoring all the denoised data P n To a denoised image, the method further includes:

[0026] Detect the edge of the tray or substrate in the denoised image, compare it with the inherent hardware size of the tray or substrate, and repair the position of the tray or substrate in the denoised image according to the comparison result to obtain an image with geometric distortion correction.

[0027] A substrate state recognition system for a chemical vapor deposition device, comprising an image capturing device, a temperature measuring device, and a processor;

[0028] The image capturing device is configured to capture an image in the reaction chamber;

[0029] The temperature measuring device is configured to collect the temperature in the reaction chamber;

[0030] The processor is communicatively connected to the image capturing device and the temperature measuring device, and is configured to implement the method steps described in any one of the above.

[0031] Optionally, the lens of the image capturing device is a telecentric lens.

[0032] Optionally, the lens of the image capturing device is an autofocus lens.

[0033] Optionally, the autofocus lens is driven by a piezoelectric ceramic driver.

[0034] A chemical vapor deposition device, comprising:

[0035] A reaction chamber, in which a rotatable tray is provided, the tray is used to carry a substrate, and an observation window is provided on the upper side of the reaction chamber;

[0036] The substrate state recognition system for the chemical vapor deposition device described in any one of the above, the image capturing device is mounted above the observation window, the viewing angle of the image capturing device is aligned with the tray, and the temperature measuring device is disposed in the reaction chamber and aligned with the tray.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] The present invention first obtains noise data corresponding to the temperatures of each node during the heating process of the reaction chamber without substrate treatment, and then, when the substrate is being processed, denoises the original images captured at the temperatures of each node according to the noise data obtained in the first stage during the heating process of the reaction chamber to obtain clear denoised images for identifying different states of the substrate. The obtained denoised images eliminate the noise caused by temperature, improve the clarity of the images, and help to accurately identify the state of the substrate. Further, according to the state of the substrate, the process parameters of the reaction chamber can be adjusted accordingly, thereby improving the quality of thin film deposition. Description of the Drawings

[0039] To more clearly illustrate the technical solution of the present invention, the accompanying drawings required for description will be briefly introduced below. Obviously, the accompanying drawings in the following description are an embodiment of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings:

[0040] Figure 1 Is a blurred image taken inside a high-temperature reaction chamber;

[0041] Figure 2a 、 2b Is a usage state diagram of a substrate state recognition system for a chemical vapor deposition device provided by the present invention;

[0042] Figure 3 Is a flowchart of a method for recognizing the state of a substrate in a chemical vapor deposition device provided by the present invention;

[0043] Figure 4a 、 Figure 4b Is a schematic diagram of the offset of the substrate relative to the disk pit;

[0044] Figure 5a 、 Figure 5b Is a schematic diagram of particle contamination on the substrate. Detailed implementation manners

[0045] The following further details the solution proposed by the present invention in conjunction with the accompanying drawings and specific implementation manners. According to the following description, the advantages and features of the present invention will be clearer. It should be noted that the accompanying drawings are in a very simplified form and all use non-precise scales, only for the purpose of facilitating and clearly assisting in explaining the embodiments of the present invention. In order to make the purpose, features, and advantages of the present invention more obvious and understandable, please refer to the accompanying drawings. It should be known that the structures, ratios, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention. Therefore, they do not have technical substantive significance. Any modification of the structure, change of the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed by the present invention.

[0046] The method, system and gas deposition equipment for identifying the substrate state involved in the present invention capture images of the tray when the reaction chamber is at the initial temperature and the temperatures of each node, compare the images corresponding to the temperatures of each node with the image corresponding to the initial temperature to obtain the noise data corresponding to the temperatures of each node. During the process of processing the substrate, during the heating process of the reaction chamber, images of the substrate and the tray when heated to any node temperature are acquired, and after being processed in combination with the noise data corresponding to this node temperature, a denoised image is obtained. The denoised image is clearer than the original image, so the accuracy of substrate state identification can be improved.

[0047] Please refer to Figure 2a - 2b , the gas deposition equipment provided by the present invention includes a reaction chamber 70. A rotatable tray 10 is provided in the reaction chamber. The tray 10 is used to carry the substrate. An observation window 60 is provided above the reaction chamber 70. The observation window usually uses transparent quartz glass, which can neither damage the vacuum degree of the chamber nor observe the local process state in the chamber. It also includes a gas deposition equipment substrate state identification system, which can be assembled above the observation window to automatically obtain information in the chamber. Figure 2a Fig. shows the usage state diagram of a gas deposition equipment substrate state identification system provided by the present invention. The gas deposition equipment substrate state identification system includes: an image capturing device 20, a temperature measuring device 30 and a processor 40. The image capturing device 20 can be a CCD (charge coupled device) industrial camera or other assembled optoelectronic devices that can record image information, and is used to capture images in the reaction chamber. The image capturing device 20 is mounted above the observation window, and the viewing angle is aligned with the tray 10 in the reaction chamber. The image capturing device 20 can capture images of the tray 10 and the substrate in the reaction chamber. In this embodiment, the temperature measuring device 30 is provided in the reaction chamber and aligned with the tray 10, and is used to collect the temperature in the reaction chamber. The temperature measuring device 30 can also measure the temperature in the reaction chamber from other positions. The specific temperature measuring device can be an infrared temperature sensor or a thermocouple contact sensor. According to the structural space in the chamber and the installation requirements of the temperature measuring device, it can be set at different positions as long as the temperature of the target component in the chamber can be obtained. The processor 40 is communicatively connected to the image capturing device 20 and the temperature measuring device 30, and is used to implement the method for identifying the substrate state of a gas deposition equipment of the present invention.

[0048] In some embodiments, the tray 10 rotates around the central axis of the tray, and at the same time, the substrate on the tray rotates around the central axis of the substrate, thereby achieving a more uniform temperature and gas flow distribution.

[0049] Such as Figure 3As shown in the figure, a method for identifying the substrate state of a chemical vapor deposition device provided by the present invention includes the following steps:

[0050] S1. Record the initial temperature T of the reaction chamber, 0 and obtain the initial image I of at least a part of the upper surface of the tray at the initial temperature T through the observation window on the reaction chamber. 0 0 ;

[0051] The initial temperature of the reaction chamber can be approximately equal to the ambient temperature or the normal temperature at the current position. At this temperature, the temperatures of other components in the reaction chamber are basically equal. The range of the initial image can be selected according to actual needs. If the process requirements only observe the local part of the substrate and can meet the prediction of the substrate state, only some areas can be selected. In other cases, the range of the obtained image can also be improved by expanding the viewing angle or continuously taking pictures and piecing them together by rotating the tray.

[0052] S2. Record n node temperatures T~T during the process of the reaction chamber rising from the initial temperature T to the final temperature, 0 and obtain the background image I of at least a part of the upper surface of the tray corresponding to each node temperature through the observation window on the reaction chamber. 1 n n ;

[0053] The node temperatures can be reasonably selected according to process conditions. For example, a preliminary estimate of the process is made, and the changes of some states on the tray or substrate after each certain temperature interval is increased, or the influence on the imaging effect. The intervals between adjacent node temperatures can be the same or different because the change of the substrate state may be non-linear. There are also multiple corresponding background images. If Tn is used as the general term of the node temperature, then In is a general term representation of the background image.

[0054] S3. The initial image I and each background image I are subjected to Fourier transform to obtain multiple frequency data and their corresponding amplitude data, and the difference between the amplitude of each background image I and the amplitude of the initial image I at the corresponding frequency is obtained to get the noise data D. 0 n n 0 n ;

[0055] ​​​​​​​Converting an image into frequency data and its corresponding amplitude data using Fourier transform belongs to common technology and will not be elaborated here. When heating up according to the node temperature without placing the substrate, the tray will spontaneously generate radiation intensity in a certain frequency band, which will mask the subtle radiation changes after placing the substrate, resulting in a blurred image of the substrate. Therefore, the radiation distribution at each node temperature in this stage is used as noise data.

[0056] S4. Place the substrate on the tray, and obtain the original image R including at least part of the upper surface of the tray and at least part of the upper surface of the substrate at each of the node temperatures. n , for each of the original images R n Subtract the amplitude data at the corresponding frequency after Fourier transform from the noise data D n to obtain the denoised data P n , where n > 0;

[0057] The data of the original image is the total radiation data during heating up after placing the substrate, which includes the radiation data of the tray heating up alone. Only the data related to the substrate is required for resolution. The radiation data of the tray will interfere with the resolution, so it is removed as noise.

[0058] S5. Use inverse Fourier transform to restore all the denoised data P n to a denoised image, and identify different states of the substrate.

[0059] Similarly, restoring frequency amplitude data to an image using inverse Fourier transform belongs to common technology and will not be elaborated here.

[0060] The method of the present invention includes two stages: steps S1 to S3 are the first stage, which is used to obtain noise data corresponding to each node temperature during the heating process of the reaction chamber without substrate processing. Steps S4 to S5 are the second stage, which is used to perform denoising processing on the original images taken at each node temperature according to the noise data obtained in the first stage during the heating process of the reaction chamber when the substrate is being processed, so as to obtain a clear denoised image for identifying different states of the substrate. The obtained denoised image eliminates the noise brought by temperature, improves the clarity of the image, and helps to accurately identify the state of the substrate. Further, according to the state of the substrate, the process parameters of the reaction chamber can be adjusted specifically, thereby improving the quality of thin film deposition.

[0061] In addition to being applied to the recognition of the substrate state during the heating process of the reaction chamber, the present invention can also be applied to the recognition of the substrate state during a period of maintaining a high temperature in the reaction chamber during the process. Since it is necessary to maintain the reaction chamber at a high temperature for a period of time after heating to the highest temperature, but the actual temperature in the reaction chamber may fluctuate continuously, and the temperatures at different positions on the tray or on the substrate may also be different, it is necessary to monitor the substrate state during this period. Specifically: record the general term T of the temperatures of i nodes of the reaction chamber from the final temperature to the end of the process n+i and the general term I of the background images at the corresponding temperatures n+i , and calculate the noise data D n+i , where i > 1; place the substrate on the tray and record the original image R n+i at the node temperature T n+i , and calculate the denoised data P n+i , obtain the denoised image, and recognize the substrate state of the substrate during the process. The method of calculating the noise data D n+i is the same as that of the noise data D n , that is, the background image I n+i is subjected to Fourier transform to obtain a plurality of frequency data and their corresponding amplitude data, and the difference between the amplitudes of the background image I n+i and the initial image I 0 at the corresponding frequencies is used to obtain the noise data D n+i . The method of calculating the denoised data P n+i is the same as that of the denoised data P n , that is, the difference between the amplitude data of the original image R n+i after Fourier transform at the corresponding frequencies and the noise data D n+i is used to obtain the denoised data P n+i , and then the inverse Fourier transform is used to restore the denoised data P n+i to the denoised image.

[0062] In the present invention, the initial temperature T 0 is the temperature before the reaction chamber is heated, usually room temperature, the final temperature is the high temperature required by the process, and each node temperature T n can be a number of temperatures evenly distributed between the initial temperature and the final temperature, or a number of landmark temperatures.

[0063] In step S4 of obtaining the original image R n , it can be that an image capturing device captures an image of part or the whole tray or substrate, or multiple image capturing devices capture images at multiple positions and then splice them to obtain the original image R n .

[0064] In this embodiment, identifying different states of the substrate includes substrate offset, the number of contamination particles, and / or substrate warping.

[0065] As Figure 2a shown, the pits on the tray 10 can be used to accommodate the substrate. To ensure the uniformity of thin film deposition, the substrate should be placed concentrically with the pits. However, during the process, the tray 10 rotates, and due to the action of centrifugal force, the substrate may be in an eccentric state with respect to the pits. Therefore, it is necessary to identify whether the substrate has shifted and the amount of shift (i.e., the size of the gap between them after the shift). In this embodiment, the denoised image obtained in step S5 may include the edges of the substrate and the pits, which is used to identify the offset amount of the substrate relative to the center of the pits. Figure 4a and Figure 4b schematically show two offset states of the substrate relative to the pits. Figure 4a In, the dark area a on the left is the tray, the arrow below points to the edge of the pit of the tray, the bright area b on the right is the substrate, the arrow above points to the edge of the substrate, and the strip area c in the middle is the gap formed after the substrate shifts to the right relative to the pit. Figure 4b In, the dark area a on the left is the tray, the bright area b on the right is the substrate, and the edge of the tray coincides with the edge of the substrate, indicating that the substrate contacts the tray after shifting to the left relative to the pit.

[0066] In some vapor deposition devices, a number of rotatable small trays 50 (such as air-bearing small trays) are provided in the pits of the tray 10. The small trays 50 are used to accommodate the substrate. During the process, the tray 10 rotates, and due to the action of centrifugal force, the small trays 50 may be in an eccentric state with respect to the pits of the tray 10. Therefore, it is necessary to identify whether the small trays have shifted and the amount of shift (i.e., the size of the gap between them after the shift). Based on this, the denoised image obtained in step S5 includes the edges of the small trays, which is used to identify the offset amount of the small trays relative to the tray.

[0067] To ensure the yield of the process products, it is necessary to monitor the particle contamination on the substrate. Figure 5a 、 5b schematically shows the particle contamination in a local area on the substrate. Thus, the denoised image obtained in step S5 includes a partial area on the substrate, which is used to judge the number of particle contaminations. At the same time, in addition to being able to judge the location where the particle contamination exists on the substrate, because a series of node temperatures correspond to different time points of the entire process, it is possible to infer at which time period of the process the particle contamination occurred based on which time node the substrate with particle contamination is in, and a change curve of the particle contamination can be plotted. It is even possible to judge during which time period the particle contamination decreases or increases.

[0068] During the process, if the substrate is warped, the captured image will be out of focus and unclear. Based on this, the lens height of the image capture device can be adjusted to capture multiple original images R ni , i>1, after processing each original image, multiple denoised images are obtained, and the lens adjustment distance corresponding to the image with the highest clarity is recorded to determine the amount of substrate warping, because the lens is above the substrate and facing a certain area on the edge of the substrate. When the substrate is warped, the lens will be blurred due to the change in focal length. Therefore, when the focal length is adjusted accordingly and refocused to obtain a clear image, the distance adjusted by the moving lens can correspond to the amount of warping. The lens of the image capture device can be an autofocus lens, in particular, it can be a fixed-focus lens with a piezoelectric ceramic driver (linear motor). By keeping the image capture device at a fixed position to observe the surface of the substrate, piezoelectric ceramics are used to control and continuously autofocus to obtain a clear image. By recording the moving distance of the linear motor, the amount of substrate warping can be determined. Of course, other types of linear motors can also be used to drive the autofocus lens.

[0069] In order to achieve comprehensive monitoring of the tray and substrate in the reaction chamber, the image capture device and the shooting position can be adjusted at any node temperature to collect original images including different positions of the upper surface of the tray and the upper surface of the substrate, thereby ensuring that different positions of the tray and the substrate can be monitored. In this embodiment, the lens of the image capture device can be a telecentric lens to meet the needs of precision detection.

[0070] Furthermore, after adjusting the blowing or temperature of the reaction chamber according to the identification results of the substrate offset, the number of contaminated particles and / or the warping state of the substrate, steps S4 and S5 can be performed again to verify whether the expected adjustment results are achieved by identifying the state of the substrate again.

[0071] In order to further improve the clarity of the denoised image, a multi-band fusion method can also be used for correction. Specifically, the image capture device can include a visible light camera and a thermal imaging camera. The two cameras simultaneously capture the same position to obtain a visible light image and an infrared image. In this embodiment, the initial image I0 includes an initial visible light image I 01 and the initial infrared image I 02 , the background image I n Including background visible light image I n1 and background infrared image I n2 , the noise data D n Including visible light image noise data D n1 and infrared image noise data D n2 . Visible light image noise data D n1 The calculation method is as follows: the initial visible light image I 01and each background visible light image I n1 Multiple frequency data and their corresponding amplitude data are obtained through Fourier transform. For each background visible light image I n1 and the initial visible light image I 01 the difference between the amplitudes at the corresponding frequencies is calculated to obtain the visible light image noise data D n1 ; the calculation method of the infrared image noise data D n2 is as follows: For the initial infrared image I 02 and each background infrared image I n2 Multiple frequency data and their corresponding amplitude data are obtained through Fourier transform. For each background infrared image I n2 and the initial infrared image I 02 the difference between the amplitudes at the corresponding frequencies is calculated to obtain the infrared image noise data D n2 .

[0072] Correspondingly, the original image R n includes the original visible light image R n1 and the original infrared image R n2 , and the denoised data P n includes the visible light denoised data P n1 and the infrared denoised data P n2 . Using the inverse Fourier transform, the visible light denoised data P n1 and the infrared denoised data P n2 are restored to a visible light image and an infrared image, and the visible light image and the infrared image are fused to obtain the denoised image. The calculation method of the visible light denoised data P n1 is as follows: For each original visible light image R n1 the amplitude data at the corresponding frequency after Fourier transform is subtracted from the visible light image noise data D n1 to obtain the visible light denoised data P n1 . The calculation method of the infrared denoised data P n2 is as follows: For each original infrared image R n1 the amplitude data at the corresponding frequency after Fourier transform is subtracted from the infrared image noise data D n2 to obtain the infrared denoised data P n2 .

[0073] In addition, since the image capturing device will inevitably produce geometric distortion when collecting images, and the inventor found that in a high-temperature environment, the degree of distortion will increase. Therefore, in this embodiment, the denoised data P nAfter restoring to the denoised image, geometric distortion correction can also be performed on the denoised image, specifically including: detecting the edges of the tray or substrate in the denoised image, comparing with the inherent hardware dimensions of the tray or substrate, and repairing the position data of the tray or substrate in the denoised image according to the comparison result to obtain the geometric distortion corrected image.

[0074] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be recognized that the above description should not be regarded as a limitation of the present invention. After those skilled in the art have read the above content, various modifications and alternatives to the present invention will be obvious. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims

1. A method for identifying the substrate state of a vapor deposition device, wherein the reaction chamber of the vapor deposition device contains a tray for carrying the substrate. Characterized in that: It includes the following steps: Record the initial temperature T of the reaction chamber 0 , and obtain an initial image I of at least a partial upper surface of the tray at the initial temperature T 0 through an observation window on the reaction chamber 0 ; Record the temperatures of n nodes T 0 during the process of increasing the temperature of the reaction chamber from the initial temperature T 1 to the final temperature, and obtain the background image I n of at least a part of the upper surface of the tray corresponding to each of the node temperatures through the observation window on the reaction chamber n ; The initial image I 0 and each background image I n are subjected to Fourier transform to obtain a plurality of frequency data and their corresponding amplitude data. The amplitude of each background image I n and the initial image I 0 at the corresponding frequency are subtracted to obtain noise data D n ; Place the substrate on the tray, and obtain an original image R including at least a partial upper surface of the tray and at least a partial upper surface of the substrate at each of the node temperatures n , and for each of the original images R n , perform a difference operation on the amplitude data at the corresponding frequency after Fourier transform and the noise data D n to obtain denoised data P n , where n > 0; Use the inverse Fourier transform to restore all denoised data P n to the denoised image and identify different states of the substrate.

2. The method for identifying the substrate state of a vapor deposition device according to claim 1. Characterized in that: The method further includes: Record the temperature T of i nodes of the reaction chamber from the final temperature to the end of the process n+1 ~T n+i and the background image I at the corresponding temperature n+i and calculate the noise data D n+i where i > 1; Place the substrate on the tray, and record the original image R at each node temperature. n+i Then, calculate the corresponding denoised data P. n+i Obtain the denoised image and identify the substrate state of the substrate during the process.

3. The method for identifying the substrate state of a vapor deposition device according to claim 1. Characterized in that: The state includes substrate offset, the number of contamination particles, and / or substrate warping.

4. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: At any node temperature, adjust the image capturing device and the capturing position to collect original images including different positions on the upper surface of the tray and the upper surface of the substrate.

5. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: The tray includes a disk pit for accommodating the substrate, and the denoised image includes the edges of the substrate and the disk pit, which are used to identify the offset of the substrate relative to the center of the disk pit.

6. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: The tray includes a rotatable small tray for accommodating the substrate, and the denoised image includes the edge of the small tray, which is used to identify the offset of the small tray relative to the large tray.

7. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: The denoised image includes a partial area on the substrate, which is used to judge the number of particle contaminations.

8. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: During the process, it also includes adjusting the lens height of the image capturing device to obtain multiple denoised images, and recording the distance of the lens adjustment corresponding to the image with the highest clarity, which is used to judge the amount of substrate warping.

9. The method for identifying the substrate state of a vapor deposition device according to claim 3. Characterized in that: The method further includes: After adjusting the gas blowing or temperature of the reaction chamber according to the recognition results of the substrate offset, the number of contaminant particles, and / or the substrate warping state, the original image R including at least a part of the upper surface of the tray and at least a part of the upper surface of the substrate at the node temperature is recorded again. n , for each of the original images R n The amplitude data at the corresponding frequency after Fourier transform and the noise data D n are subtracted to obtain the denoised data P n , and the denoised data P is restored to the denoised image by using the inverse Fourier transform n in the step of identifying different states of the substrate.

10. The method for identifying the substrate state of a vapor deposition device according to claim 1. Characterized in that: The initial image I 0 includes an initial visible light image I 01 and an initial infrared image I 02 , and the background image I n includes a background visible light image I n1 and a background infrared image I n2 , and the noise data D n includes visible light image noise data D n1 and infrared image noise data D n2 ; the original image R n includes the original visible light image R n1 and the original infrared image R n2 The denoised data P n includes the visible light denoised data P n1 and the infrared denoised data P n2 The visible light denoised data P n1 and the infrared denoised data P n2 are restored to a visible light image and an infrared image by using inverse Fourier transform, and the visible light image and the infrared image are image - fused to obtain the denoised image.

11. The method for identifying the substrate state of a vapor deposition device according to claim 1. Characterized in that: After restoring all the denoised data P n to the denoised image, the method further includes: Detect the edge of the tray or the substrate in the denoised image, compare it with the inherent hardware size of the tray or the substrate, and repair the position of the tray or the substrate in the denoised image according to the comparison result to obtain a geometric distortion corrected image.

12. A substrate state identification system for a vapor deposition device. Characterized in that: It includes an image capturing device, a temperature measuring device, and a processor; The image capturing device is used to capture images in the reaction chamber; The temperature measuring device is used to collect the temperature in the reaction chamber; The processor is communicatively connected to the image capturing device and the temperature measuring device, and is used to implement the method steps according to any one of claims 1-11.

13. The substrate state identification system for a vapor deposition device according to claim 12. Characterized in that: The lens of the image capturing device is a telecentric lens.

14. The substrate state identification system for a vapor deposition device according to claim 12. Characterized in that: The lens of the image capturing device is an autofocus lens.

15. The substrate state recognition system of the vapor deposition equipment according to claim 12, characterized in that, the autofocus lens is driven by a piezoelectric ceramic driver.

16. A vapor deposition equipment, characterized in that, comprising: a reaction chamber, a rotatable tray is arranged in the reaction chamber, the tray is used for carrying a substrate, and an observation window is arranged on the upper side of the reaction chamber; the substrate state recognition system of the vapor deposition equipment according to any one of claims 12-15, the image capturing device is mounted above the observation window, the viewing angle of the image capturing device is aligned with the tray, and the temperature measuring device is arranged in the reaction chamber and aligned with the tray.

Images