Substrate abnormality detection method and substrate abnormality detection apparatus

Abstract

The present disclosure relates to a substrate abnormality detection method and a substrate abnormality detection apparatus that accurately detect a crack or the like of a substrate. A control section of the substrate abnormality detection apparatus acquires a moving image including a state in which a substrate is separated from a placement surface of a placement table by a lift pin, captured by a camera that captures an inside of a processing container via an observation window provided to a side wall of the processing container, sets an evaluation target range to a portion corresponding to the placement surface, so that an image of a pin-up state in which the substrate is separated from the placement surface among a plurality of images constituting the moving image does not include the substrate, performs adaptive binarization processing on a gray scale image that represents the image of the pin-up state using a gray scale, and determines that an abnormality has occurred in the substrate in a case where a proportion of pixels of white in a binarized image acquired by performing the adaptive binarization processing on the evaluation target range is equal to or greater than a threshold value decided in advance.

Description

Technical Field

[0001] This disclosure relates to a substrate anomaly detection method and a substrate anomaly detection device. Background Technology

[0002] A plasma processing apparatus is known to perform plasma processing on a substrate placed on a stage inside a processing container. In such a plasma processing apparatus, anomalies, such as cracks, in the plasma-treated substrate are detected. For example, in the substrate anomaly detection method of Patent Document 1, cracks in the substrate are detected by a sensor located near a gate valve connecting the processing container and the conveying device during the removal of the plasma-treated substrate from the processing container and its transfer to a conveying device.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2007-73599 Summary of the Invention

[0006] The problem the invention aims to solve

[0007] The technology disclosed herein accurately detects abnormalities such as cracks in substrates.

[0008] Solution for solving the problem

[0009] One aspect of the technology disclosed herein is a substrate anomaly detection method, which is a method for detecting anomalies in a substrate being processed inside a processing container of a substrate processing apparatus. The substrate processing apparatus includes: a stage disposed inside the processing container and having a mounting surface for mounting the substrate; a lifting pin capable of protruding from and engaging with the mounting surface to raise or lower the substrate; at least one observation window disposed on a side wall of the processing container; and a control device for controlling the execution of the processing performed on the substrate. The substrate anomaly detection method includes the following steps: at least obtaining data obtained from observations of the interior of the processing container via the observation window. The system captures a motion image captured by at least one camera, including a state in which the substrate is detached from the mounting surface by the lifting pin; it sets an evaluation object range for the portion corresponding to the mounting surface such that the substrate is not included in the image of the pin-raised state in which the substrate is detached from the mounting surface among a plurality of images constituting the motion image; it performs adaptive binarization processing on the evaluation object range of the grayscale image representing the pin-raised state using grayscale; and it determines that the substrate has an abnormality if the proportion of white pixels in the binarized image obtained by performing adaptive binarization processing on the evaluation object range is above a predetermined threshold.

[0010] The effects of the invention

[0011] According to the technology disclosed herein, it is possible to accurately detect anomalies such as cracks in a substrate. Attached Figure Description

[0012] Figure 1 This is a block diagram schematically illustrating the structure of a substrate anomaly detection system, which includes an information processing apparatus for substrate anomaly detection as one embodiment of the technology involved in this disclosure.

[0013] Figure 2 It is shown in a general way. Figure 1 A cross-sectional view of the structure of the film-forming device.

[0014] Figure 3 It is used to explain by Figure 2 An image of the interior of the processing container taken by a camera.

[0015] Figure 4 It is shown in a general way. Figure 1 A block diagram of the hardware structure of an information processing device.

[0016] Figure 5 It is a diagram used to illustrate an image obtained by projecting fragments of the substrate, the shadow of the substrate, and the rising state of the shooting pin of the shielding mesh.

[0017] Figure 6 This is a diagram used to illustrate adaptive binarization processing.

[0018] Figure 7 This diagram illustrates the application of adaptive binarization to a grayscale image in an ascending state.

[0019] Figure 8 It is a binary image used to illustrate the scope of the evaluation object.

[0020] Figure 9 It is shown as by Figure 1 The flowchart describes the substrate anomaly detection process executed by the information processing device.

[0021] Figure 10 It is used to explain becoming Figure 9 The image is a picture of the substrate abnormality detection processing object with the pin in the rising state.

[0022] Figure 11 It is used for explanation Figure 10 The image is a binary image of the evaluation object range in the image.

[0023] Figure 12This is a diagram used to illustrate the morphological processing performed on a binarized image of the evaluation object range.

[0024] Figure 13 This diagram illustrates the erosion and dilation processes within the opening and closing operations of morphological processing. Detailed Implementation

[0025] Hereinafter, an embodiment of the technology involved in this disclosure will be described with reference to the accompanying drawings. Figure 1 This is a block diagram schematically illustrating the structure of a substrate anomaly detection system 1, including an information processing apparatus 40, which is an embodiment of the technology involved in this disclosure. Figure 1 In this system, the substrate anomaly detection system 1 comprises a film deposition apparatus 10 (which serves as a substrate processing apparatus), a camera 22, and an information processing unit 40. In the substrate anomaly detection system 1, the information processing unit 40 performs the following steps on the moving images transmitted from the camera 22: (described later) Figure 9 Substrate anomaly detection and processing, determining the placement stage 13 inside the processing container of the film deposition apparatus 10 (refer to...) Figure 2 The film forming apparatus 10 is notified whether the substrate placed on it has cracked.

[0026] Figure 2 It is shown in a general way. Figure 1 A cross-sectional view of the structure of the film-forming device 10. Figure 2 The film deposition apparatus 10 is an inductively coupled plasma processing apparatus. The film deposition apparatus 10 uses plasma generated from the processing gas to perform film deposition processing such as forming a protective film on a rectangular substrate, such as a glass substrate G (hereinafter referred to as "substrate G") for an FPD (Flat Panel Display).

[0027] The film-forming apparatus 10 includes a cylindrical processing container 11 made of a conductive material, which houses a substrate G. The upper part of the processing container 11 is hermetically sealed by a window member 12. Inside the processing container 11, a mounting stage 13 for placing the substrate G is disposed at the lower part, facing the window member 12. Furthermore, a processing space U is formed between the mounting stage 13 and the window member 12 within the processing container 11. In the processing space U, as described later, plasma is generated from a processing gas.

[0028] An electrostatic holding disk (not shown) is provided on the upper surface of the mounting stage 13. The substrate G placed on the mounting stage 13 is held on the mounting stage 13 by the electrostatic holding disk to which a DC voltage is applied. In addition, a lifting pin 38 (lifting pin) is provided on the mounting stage 13 (see reference). Figure 3 (B)

[0029] The lifting pin 38 protrudes from and inserts into the mounting surface of the mounting table 13, enabling the substrate G to be raised and lowered. The lifting pin 38 is primarily used when moving the substrate G between the inside and outside of the processing container 11 via the loading / unloading outlet 19 (described later). More specifically, the lifting pin 38 lifts the substrate G, detaching it from the mounting table 13, and transfers the substrate G between a loading / unloading device (not shown) located adjacent to and entering the interior of the processing container 11. Furthermore, the detachment of the substrate G from the mounting table 13 by the lifting pin 38 will henceforth be referred to as "pin lifting." Additionally, the lifting pin 38 is also used when performing a destatic treatment on the substrate G that is electrostatically adsorbed onto the mounting table 13.

[0030] In the static elimination process of substrate G, firstly, the DC voltage applied to the electrostatic holding disk is stopped after the film formation process is completed. Next, a static elimination gas, such as argon (Ar), nitrogen (N2), or oxygen (O2), is supplied to the interior of the processing container 11 to generate static elimination plasma. Then, after generating the static elimination plasma, the substrate G is raised by the lifting pin 38 to improve the static elimination effect.

[0031] Furthermore, inside the mounting stage 13, a temperature control mechanism such as a cooler and a heat transfer gas supply mechanism (neither shown) are provided to control the temperature of the mounted substrate G. The temperature control mechanism adjusts the temperature of the mounting stage 13, and the heat transfer gas supply mechanism supplies heat transfer gas between the back surface of the substrate G and the mounting surface of the mounting stage 13. In addition, the mounting stage 13 is mounted on the bottom surface of the processing container 11 via an insulating frame 14.

[0032] A metal frame 15 is provided at the upper end of the side wall of the processing container 11, and a side wall portion 16 is provided on the upper surface of the metal frame 15. The side wall portion 16 supports the top plate 17, and the top plate 17 covers the window member 12 from above. An O-ring or other sealing member 18 is provided between the side wall of the processing container 11 and the metal frame 15 to keep the processing space U airtight.

[0033] A loading / unloading outlet 19 for moving the substrate G into and out of the processing space U, and a gate valve 20 for opening and closing the loading / unloading outlet 19 are provided on one side wall of the processing container 11. Additionally, a monitoring window 21 (observation window) made of an optically transparent component is provided on the other side wall of the processing container 11 and is hermetically embedded therein. A camera 22 (video camera) is provided outside the processing container 11 to capture color images of the interior of the processing container 11 through the monitoring window 21. The camera 22 can be provided as part of the film-forming apparatus 10, or it can be provided separately from the film-forming apparatus 10. Furthermore, a shielding net 39 is provided on the side of the monitoring window 21 facing inwards from the processing container 11. The shielding net 39 blocks the plasma generated in the processing space U to prevent the plasma from reaching the monitoring window 21. The camera 22 transmits the captured images of the interior of the processing container 11 to the information processing device 40. Furthermore, in this embodiment, the images of the interior of the processing container 11 captured by the camera 22 are moving images, but they can also be still images. In addition, the motion images captured by camera 22 include a series of information related to the mounting surface of the mounting stage 13 from the moment the substrate G is placed onto the mounting stage 13 until the substrate G is lifted by the lifting pin 38.

[0034] The window member 12 is rectangular and divided into multiple segments 23 made of a conductor, such as metal. Separating members 24 made of an insulator are disposed between each adjacent segment 23. The separating members 24 separate adjacent segments 23 and electrically insulate them from each other.

[0035] In the film-forming apparatus 10, the space enclosed by the window member 12, the side wall portion 16, and the top plate 17 constitutes the antenna chamber 25. In the antenna chamber 25, an inductively coupled antenna 26 is arranged facing the substrate G of the mounting stage 13 through the window member 12. Furthermore, the inductively coupled antenna 26 is arranged as a whole across the area facing each segment 23, and is formed in a vortex shape that surrounds the circumference of the window member 12, but the shape of the inductively coupled antenna 26 is not limited to a vortex shape.

[0036] Each segment 23 has a large number of vents 27 opening toward the processing space U. A gas supply pipe 28 is connected to each segment 23, and the gas supply pipe 28 is connected to a gas supply device 29. Furthermore, a gas diffusion chamber 30 is formed inside each segment 23, and the gas supply device 29 introduces processing gas into the gas diffusion chamber 30 via the gas supply pipe 28. The processing gas introduced into the gas diffusion chamber 30 is supplied to the processing space U through each vent 27.

[0037] In the film-forming apparatus 10, the inductively coupled antenna 26 is connected to the high-frequency power supply 32 via a matching adapter 31. The high-frequency power supply 32 supplies, for example, 13.56 MHz of high-frequency power for plasma generation to the inductively coupled antenna 26. As a result, eddy currents are induced in each segment 23 constituting the window member 12, circulating from the upper surface (towards the inductively coupled antenna 26) to the lower surface (towards the processing space U), and an induced electric field is formed in the processing space U by these eddy currents. Furthermore, this induced electric field excites the processing gas supplied to the processing space U to generate plasma. In addition, in this embodiment, the window member 12 is described as being made of a conductive material such as metal, but the window member 12 may also be made of a dielectric. In this case, the electric field formed by the inductively coupled antenna 26 passes through the window member 12 and acts directly on the processing gas to generate plasma. In addition, gas diffusion chambers, vents, etc., are formed, for example, on the support beam supporting the dielectric.

[0038] Furthermore, the stage 13 is connected to the high-frequency power supply 34 via a matching adapter 33. The high-frequency power supply 34 supplies, for example, 3.2 MHz of bias power to the stage 13. This attracts various ions from the plasma in the processing space U to the substrate G, performing a film deposition process to form various films on the substrate G. Alternatively, in the film deposition apparatus 10, etching can also be performed on the substrate G by attracting various ions to it. In this case, the film deposition apparatus 10 functions as an etching apparatus.

[0039] Furthermore, in the film-forming apparatus 10, an exhaust port 35 is formed on the bottom surface of the processing container 11. An exhaust device 36, such as a turbomolecular pump or a dry pump, is connected to this exhaust port 35. During the film-forming process, the exhaust device 36 maintains the processing space U at a predetermined pressure below atmospheric pressure. Additionally, a control unit 37 is provided in the film-forming apparatus 10. The control unit 37 is a control device composed of a computer having at least a CPU and memory, and records in the memory a process (program) for performing a series of substrate processing operations, such as film-forming operations and static electricity removal on the substrate G. The process also includes information for executing pin rise on the substrate G. The control unit 37 controls the execution of various processes performed on the substrate G.

[0040] Figure 3 It is used to explain by Figure 2 A diagram showing an image of the interior of the processing container 11 taken by camera 22. Figure 3 (A) and Figure 3 (B) shows one of the multiple images that constitute a moving image captured by camera 22. In addition, to be precise, a moving image consists of multiple frames, each of which includes the captured image (photographic image), but for the sake of simplicity, the concept of "frame" will not be used below, and the premise is that a moving image consists of multiple images.

[0041] As described above, the camera 22 captures images of the interior of the processing container 11 via the monitoring window 21. Specifically, it captures images from the state where the substrate G is placed on the mounting stage 13 until the state where the substrate G is pinned up by the lifting pin 38 (hereinafter referred to as the "pin-raised state"), obtaining images including... Figure 3 The photographic image of (A), Figure 3 (B) is a moving image of a photographic image. It was obtained with the photographic substrate G placed on the mounting stage 13. Figure 3 The photographic image (A) includes the internal structural components of the processing container 11, such as the mounting stage 13 and the substrate G placed on the mounting stage 13. Additionally, the image obtained in the photographic pin-up position... Figure 3 The photographic image (B) includes the stage 13, the substrate G, and the lifting pin 38 that lifts the substrate G. Then, the camera 22 sends the captured moving image to the information processing device 40. In addition, in this embodiment, the camera 22 sends the moving image directly to the information processing device 40, but it is also possible for the camera 22 to send the moving image to the information processing device 40 via the control unit 37 of the film forming apparatus 10.

[0042] Figure 4 It is shown in a general way. Figure 1 The diagram shows the hardware structure of the information processing device 40. The information processing device 40 is, for example, a personal computer. The information processing device 40 includes a CPU 41 (control unit), ROM 42, RAM 43, HDD 44, communication I / F 45, and input / output I / F 46, which are connected to each other in a communicable manner via a system bus 47.

[0043] CPU 41 is a microprocessor that unifies the control of information processing device 40. Specifically, CPU 41 reads programs stored in ROM 42 and HDD 44, and executes programs deployed in RAM 43. ROM 42 stores programs executed by CPU 41, etc. RAM 43 is used as the main memory of CPU 41, and is also used as a temporary storage area for programs and data. HDD 44 stores programs executed by CPU 41 and various data. Furthermore, in this embodiment, the structure of information processing device 40 having an HDD as an example of a storage device is described, but it is also possible for information processing device 40 to have a storage device other than an HDD, such as an SSD.

[0044] Communication I / F 45 is an interface for communication between information processing device 40 and external devices. For example, communication I / F 45 receives moving images captured by camera 22. Furthermore, communication via communication I / F 45 can be wireless or wired, implemented using various communication technologies. Input / output I / F 46 connects to output devices such as displays showing various information, and input devices for user use such as mice and keyboards.

[0045] Furthermore, in the substrate anomaly detection method described in Patent Document 1, cracks in the substrate G are detected by a sensor (not shown) located near the gate valve 20 connecting the processing container 11 to the conveying device (not shown). Therefore, even if a crack in the substrate G occurs during the film deposition process inside the processing container 11, the anomaly, such as the crack in the substrate G, cannot be detected before the substrate G is removed from the processing container 11. As a result, fragments of the cracked substrate G are scattered not only on the loading stage 13 of the processing container 11 but also near the gate valve 20, potentially extending the downtime for removing the fragments.

[0046] Therefore, it is considered to detect early on whether the substrate G has cracked based on a moving image obtained by capturing the interior of the processing container 11, including the mounting stage 13. In this case, the CPU 41 of the information processing device 40 obtains a grayscale image by acquiring an image representing the pin-up state in grayscale from the moving image captured by the camera 22. Furthermore, the CPU 41 binarizes the grayscale image of the pin-up state to obtain a binarized image. At this time, the fragments of the substrate G remaining on the mounting surface of the mounting stage 13 are transformed into white. Moreover, if the proportion of white pixels in the binarized image is above a predetermined threshold, the CPU 41 determines that there are fragments 49 of the substrate G remaining on the mounting surface of the mounting stage 13, that is, an abnormality such as a crack in the substrate G.

[0047] However, in images obtained by shooting the pin in its raised state, such as Figure 5 As shown, sometimes not only fragments 49 of the substrate G are reflected, but also the shadow 50 of the substrate G generated by the emission of plasma, and the shielding net 39 placed in front of the camera that takes pictures of the inside of the processing container 11. Furthermore, in this case, since the camera 22 is not focused on the shielding net 39, therefore, as... Figure 5 As shown, in the image obtained when the shooting pin is in the raised state, the shading net 39 is reflected as a blurry shadow.

[0048] Furthermore, when converting pixels with values ​​below a threshold to white during the binarization of a grayscale image in the pin-up state, the shadow 50 and the masking mesh 39 of the substrate G may sometimes be converted to white as well. Therefore, if the shadow 50 and the masking mesh 39 of the substrate G are reflected in the image obtained by capturing the image on the mounting stage 13, even if no fragments 49 of the substrate G remain on the mounting surface of the mounting stage 13, it may sometimes be judged that the substrate G has cracked, making it difficult to accurately detect abnormalities such as cracks in the substrate G. Additionally, in the grayscale image in the pin-up state, the small fragments 49 and the masking mesh 39 are almost the same size; therefore, it is impossible to distinguish between the small fragments 49 and the masking mesh 39 that have been converted to white in the binarized image. As a result, it is impossible to detect the small fragments 49 remaining on the mounting stage 13.

[0049] In response, the technology disclosed herein performs adaptive binarization processing on the grayscale image obtained by representing the pin rising state in grayscale based on the motion image captured by the camera 22. Instead of extracting the entire fragment 49 of the substrate G, it extracts its outline and prevents the shadow 50 of the substrate G and the occlusion mesh 39 of the substrate G in the binarized image from being transformed into white.

[0050] Figure 6 This is a diagram used to illustrate adaptive binarization processing. Figure 6 In the example shown, a grayscale image consisting of 5×6 pixels is presented. 100, 120, 140, 160, 180, and 200 represent the pixel values ​​of each pixel. Pixels shown at the same density have the same pixel value.

[0051] Adaptive binarization is performed as follows: In an image, multiple pixels are defined that enclose the pixel that forms the object (hereinafter referred to as the "center pixel"). Figure 6 In the binarization calculation region (the eight pixels adjacent to the central pixel in the vertical, horizontal, and diagonal directions), the average pixel value of each pixel in the binarization calculation region is used as the binarization threshold to binarize the pixel value of the central pixel. Specifically, if the pixel value of the central pixel is above the binarization threshold, the pixel value of the central pixel is transformed to 0 (black); if the pixel value of the central pixel is below the binarization threshold, the pixel value of the central pixel is transformed to 255 (white).

[0052] Here, as methods for calculating the binarization threshold, examples include calculating the average value based on the pixel values ​​of all pixels in the binarization calculation region, including the central pixel, and calculating the average value based on the pixel values ​​of each pixel in the binarization calculation region excluding the central pixel. Below, the case of calculating the binarization threshold using the latter method will be explained. For example, in... Figure 6In example (A), for the central pixel A, which is surrounded only by pixels with a pixel value of 200, the binarization threshold is 200, so the central pixel A is transformed into black. For the central pixel B, which is surrounded by pixels with a pixel value of 200 and pixels with a pixel value of 100, the binarization threshold is 162.5, so the central pixel B is transformed into black. Similarly, for the central pixel D, which is surrounded only by pixels with a pixel value of 100, the binarization threshold is 100, so the central pixel D is transformed into black. On the other hand, for the central pixel C, which is surrounded by pixels with a pixel value of 200 and pixels with a pixel value of 100, the binarization threshold is 137.5, so the central pixel C is transformed into white.

[0053] Additionally, for example, in Figure 6 In example (B), regarding the central pixel A with a pixel value of 180, which is surrounded by pixels with a pixel value of 200, 180, and 160, the binarization threshold is 180, therefore the central pixel A is transformed into black. Regarding the central pixel B with a pixel value of 160, which is surrounded by pixels with a pixel value of 180, 160, and 140, the binarization threshold is 160, therefore the central pixel B is transformed into black. Regarding the central pixel C with a pixel value of 140, which is surrounded by pixels with a pixel value of 160, 140, and 120, the binarization threshold is 140, therefore the central pixel C is transformed into black. Regarding the central pixel D with a pixel value of 120, which is surrounded by pixels with a pixel value of 140, 120, and 100, the binarization threshold is 120, therefore the central pixel D is transformed into black.

[0054] That is, when adaptive binarization is applied to a grayscale image, the central pixel in the binarization calculation region whose pixel value suddenly becomes small relative to any of the pixels surrounding the central pixel is transformed into white. On the other hand, the central pixels in the binarization calculation region whose pixel values ​​do not suddenly become small relative to any of the pixels surrounding the central pixel, and the central pixels whose pixel values ​​do not change abruptly but gradually, are transformed into black.

[0055] In addition, such as Figure 5 As shown, in the image obtained when the pin is in the raised state, the brightness and saturation change abruptly at the boundary between the fragment 49 of the substrate G and the stage 13. Therefore, in the grayscale image of the raised state, the pixel value changes abruptly at the boundary between the fragment 49 of the substrate G and the stage 13. On the other hand, as... Figure 5As shown, in the image obtained when the pin is in the raised state, the brightness and saturation at the shadow 50 and the masking mesh 39 of the substrate G only change gradually. Therefore, in the grayscale image of the raised pin state, the pixel values ​​at the shadow 50 and the masking mesh 39 of the substrate G do not change abruptly but only gradually. Thus, when adaptive binarization processing is performed on the grayscale image of the raised pin state, the boundary between the fragment 49 of the substrate G and the stage 13 is transformed into white, while the shadow 50 and the masking mesh 39 of the substrate G are transformed into black. Therefore, in this embodiment, the boundary between the fragment 49 of the substrate G and the stage 13, i.e., the outline of the fragment 49 of the substrate G, is transformed into white from the grayscale image of the raised pin state, and the shadow 50 and the masking mesh 39 of the substrate G are prevented from being transformed into white.

[0056] Figure 7 This diagram illustrates the application of adaptive binarization to a grayscale image in an ascending state. Furthermore, Figure 7 The explanation is based on the premise that no fragments 49 of the substrate G remain on the mounting surface of the mounting stage 13, but the shadow 50 of the substrate G and the shielding net 39 are reflected in the image obtained by photographing the mounting stage 13.

[0057] First, in this embodiment, in the grayscale image of the pin-raised state, the portion corresponding to the mounting surface of the mounting stage 13 is defined as the evaluation target range. In setting the evaluation target range, the coordinates 51a to 51d of four points are specified to enclose the portion corresponding to the mounting surface of the mounting stage 13. At this time, the user specifies the coordinates in a manner that the pin-raised substrate G is not included in the evaluation target range. Figure 7 (A)).

[0058] Next, adaptive binarization processing is performed on the evaluation object range. At this time, for example, in part A of the shadow 50 partially projected onto the substrate G and the occlusion mesh 39, regarding the central pixel with a pixel value of 100 in the binarization calculation region containing pixels with a pixel value of 70, a pixel with a pixel value of 100, and a pixel with a pixel value of 120, the binarization threshold is 97.5. Therefore, this central pixel is transformed into black (…). Figure 7 (B)). In this way, within the evaluation area, the central pixel of the portion including the shadow 50 of the substrate G and the shading mesh 39 is transformed to black, therefore... Figure 8 As shown in (A), in the binarized image of the evaluation object range after adaptive binarization processing, the shadow 50 of the substrate G and the occlusion mesh 39 are not transformed into white. Furthermore, as... Figure 7 As shown in (A), the evaluation object area includes the lifting pin 38, and at the boundary between the lifting pin 38 and the mounting stage 13, the pixel value also changes abruptly. Therefore, in the binarized image of the evaluation object area after adaptive binarization processing, as shown... Figure 8As shown in (A), the outline of the lifting pin 38 is changed to white.

[0059] On the other hand, when the evaluation object range is subjected to conventional binarization instead of adaptive binarization, if the pixel values ​​of the shadow 50 of the substrate G and the occlusion net 39 in the grayscale image are below the binarization threshold, such as... Figure 8 As shown in (B), not only the lifting pin 38, but also the shadow 50 of the substrate G and the shielding net 39 are changed to white.

[0060] Furthermore, in the adaptive binarization process, as described above, when the average pixel value of the multiple pixels surrounding the central pixel is simply set as the binarization threshold, the pixel values ​​of the shadow 50 of the substrate G and a portion of the occlusion mesh 39 change significantly, even if not drastically. Therefore, sometimes these parts of the shadow 50 of the substrate G and the occlusion mesh 39 are transformed into white. In such cases, the binarization threshold is adjusted by adjusting the value C, as shown in equation (1) below.

[0061] Threshold = Average of the pixel values ​​of the multiple pixels surrounding the central pixel - C…(1)

[0062] At this point, the user repeatedly performs adaptive binarization processing on the evaluation object range while changing the adjustment value C, setting the value where the shadow 50 of the substrate G and the shading mesh 39 are all transformed into black as the adjustment value C. Alternatively, the user can also repeatedly perform adaptive binarization processing on the evaluation object range while changing the adjustment value C, setting the value where the outline of the lifting pin 38 is clearly transformed into white as the adjustment value C.

[0063] Alternatively, the setting of the adjustment value C can be performed not by repeating the adaptive binarization process performed by the user each time, but by using machine learning based on learning data pre-generated through repeated adaptive binarization processes performed by the user. In this case, a grayscale image of the evaluation object range is input into a learning-completed model configured as a convolutional neural network, and the adjustment value C is output from the learning-completed model. The learning-completed model is configured to be generated by machine learning using the learning data, and outputs the adjustment value C when a grayscale image of the evaluation object range is input. The learning data here is teacher data used for supervised learning, and consists of a large number of grayscale images of the evaluation object range used for learning, including the images of the shadow 50 of the substrate G and the occlusion net 39, and the appropriate adjustment value C corresponding to each of these grayscale images (the adjustment value C where the images of the shadow 50 of the substrate G, the occlusion net 39, etc., are all transformed into black). The appropriate adjustment value C here can be obtained by deriving a binarized image from the grayscale image of the object that becomes the teacher data without using the adjustment value, and then adjusting the adjustment value C in such a way as deriving a binarized image of the image without the shadow 50 of the substrate G, the occlusion net 39, etc. from the grayscale image. Moreover, a large number of such combinations of adjustment value C and grayscale images can be prepared and used as teacher data.

[0064] Furthermore, the binarization threshold can be adjusted by changing the calculation method instead of adjusting the value C. For example, in the above embodiment, the binarization calculation region is formed by multiple pixels surrounding the central pixel in a single layer, and the average pixel value of each pixel included in the binarization calculation region is used as the binarization threshold. However, the binarization calculation region can also be formed by multiple pixels surrounding the central pixel in two layers, and the average pixel value of each pixel included in the binarization calculation region is used as the binarization threshold. In addition, as needed, the multiple pixels surrounding the central pixel can also be multiple pixels surrounding the central pixel in three or more layers. Furthermore, in the above embodiment, in the binarization calculation region, the average pixel value of only the multiple pixels surrounding the central pixel is used as the binarization threshold (that is, the pixel value of the central pixel is not reflected in the binarization threshold). However, in the binarization calculation region, the average pixel value of the multiple pixels surrounding the central pixel and the pixel value of the central pixel can also be used as the binarization threshold.

[0065] Figure 9 It is shown as by Figure 1 The flowchart of the substrate anomaly detection process executed by the information processing device 40 is shown. Figure 9 The substrate anomaly detection and processing is implemented by CPU 41 executing the program stored in ROM 42 and HDD 44. Figure 9The substrate anomaly detection process is performed, for example, when the information processing device 40 receives a start instruction for substrate anomaly detection processing performed by the user. Furthermore, in this embodiment, the process used during a series of processes performed by the film deposition apparatus 10, such as film deposition and static electricity removal from the substrate G, is pre-saved in the storage device (built-in or external) of the control unit 37. The process saved in the storage device is executed in the memory of the control unit 37 during processing. This process includes, for example, the end time of the pin rise of the substrate G.

[0066] exist Figure 9 First, the CPU 41 acquires a moving image captured by the camera 22 (S91). As described above, this moving image includes an image of the pin rising state, which is the object of anomaly detection, i.e., an image showing the state in which the substrate G is detached from the mounting surface. Additionally, the CPU 41 acquires the process data stored in the storage device of the control unit 37. At this time, it is not necessary to acquire all the information included in the process data to the CPU 41; only information related to the end time of the pin rising is required. Furthermore, the timing of acquiring the process data is not limited to this timing; the information processing device 40 may also acquire the process data from the control unit 37 in advance before starting this process. Furthermore, in S91, the moving image may also be associated with the time included in the process data.

[0067] Next, the CPU 41 uses the coordinates 51a to 51d of four points pre-specified by the user to set the evaluation object range for the acquired motion image (S92). Furthermore, in this embodiment, before starting this process, the user uses an image obtained by capturing the pin in the raised state (the image of the pin in the raised state), and specifies the coordinates 51a to 51d of four points to enclose the portion of the image corresponding to the mounting surface of the mounting platform 13 (see reference). Figure 7 (A)).

[0068] Next, based on the acquired motion image, the CPU 41 converts the image obtained from capturing the pin rising state to grayscale to obtain (generate) a grayscale image of the pin rising state (S93). Specifically, the CPU 41 acquires an image from the acquired motion image that corresponds to the moment (e.g., the end time of pin rising) representing the pin rising state of the substrate G included in the process, and converts that image to grayscale to generate a grayscale image of the pin rising state.

[0069] Next, CPU 41 performs adaptive binarization processing (S94) on the evaluation object range of the grayscale image in the rising state. Specifically, all pixels surrounding a pixel in the evaluation object range are taken as the center pixels, and a binarization calculation region with multiple pixels surrounding each center pixel is set. Adaptive binarization processing is performed on all binarization calculation regions to obtain the binarized image of the evaluation object range. That is, each pixel surrounding a pixel in the evaluation object range is a center pixel, and also a pixel constituting the binarization calculation region of the other center pixels. At this time, even if... Figure 10 As shown, the image obtained in the rising position of the shooting pin not only reflects the fragment 49 of the substrate G, but also the shadow 50 of the substrate G and the shielding mesh 39, and so on. Figure 11 As shown, in the binarized image of the evaluation object range after adaptive binarization processing, the shadow 50 of the substrate G and the occlusion mesh 39 are not transformed into white, only the outlines of the lifting pin 38 and the fragments 49 of the substrate G are transformed into white.

[0070] Next, the CPU 41 determines whether the proportion of white pixels in the binarized image of the evaluation object area is above a predetermined anomaly detection threshold (S95). Here, the anomaly detection threshold varies depending on the type of substrate G. For example, when substrate G is a patterned substrate, sometimes not only the outlines of the lifting pin 38 and the fragments 49 of substrate G, but also the outlines of the pattern are transformed into white. Therefore, the proportion of white pixels in the binarized image of the evaluation object area is larger than that of substrate G made of plain glass without a pattern. Therefore, the anomaly detection threshold for patterned substrates is set to be larger than that for substrate G made of plain glass. For example, the anomaly detection threshold for patterned substrates is set to 4%, and the anomaly detection threshold for substrate G made of plain glass is set to 2%. Furthermore, the anomaly detection thresholds for patterned substrates and substrate G made of plain glass are not limited to these values.

[0071] Furthermore, the threshold for anomaly detection is set by the user through prior experiments, but the threshold can also be set using machine learning. In this case, a binarized image of the evaluation object range can be input into a learning-completed model configured as a convolutional neural network, and the anomaly detection threshold can be output from the learning-completed model. The learning-completed model is configured to be generated using machine learning with learning data, and outputs the anomaly detection threshold when a binarized image of the evaluation object range is input. Here, the learning data includes a large number of binarized images of the evaluation object range containing white pixels representing the outline of fragment 49 of the substrate G and the lifting pin 38, and these images are associated with appropriate anomaly detection thresholds corresponding to the binarized images of each evaluation object range.

[0072] If, in S95, the proportion of white pixels in the binarized image of the evaluation target area exceeds a threshold, then the outline of the substrate G fragment 49 clearly appears in the binarized image of the evaluation target area. Therefore, the CPU 41 notifies the film deposition apparatus 10 that the substrate G has developed an abnormality such as a crack (S96), and terminates the process. Upon receiving this notification, the film deposition apparatus 10 stops the ongoing power-off process to prevent the expansion of damage caused by the abnormality of the substrate G.

[0073] On the other hand, in S95, if the proportion of white pixels in the binarized image within the evaluation target range is less than a threshold, the CPU 41 notifies the film deposition apparatus 10 that the substrate G has not generated any abnormality (S97), and the process ends. Upon receiving this notification, the film deposition apparatus 10 continues the ongoing power-removal process. Alternatively, it can be configured such that if the proportion of white pixels in the binarized image within the evaluation target range is less than a threshold, no special notification is sent to the film deposition apparatus 10, and the film deposition apparatus 10 continues the ongoing power-removal process as usual. Furthermore, in this process, from the acquisition of the motion image in S91, the notification in S96 or S97 is performed within a relatively short time, for example, about 1 second.

[0074] According to this embodiment, adaptive binarization processing is performed on the evaluation object range of the grayscale image in the pin-up state. In the adaptive binarization processing, the central pixel whose pixel value abruptly decreases to a small value relative to the surrounding pixels is transformed into white; the central pixels whose pixel values ​​remain unchanged relative to the surrounding pixels, and the central pixels whose pixel values ​​only gradually change relative to the surrounding pixels, are transformed into black. Furthermore, in the grayscale image in the pin-up state, the pixel value abruptly changes at the boundary between the fragment 49 of the substrate G and the mounting stage 13; on the other hand, the pixel values ​​only gradually change at the shadow 50 and the occlusion mesh 39 of the substrate G. Therefore, when adaptive binarization processing is performed on the grayscale image in the pin-up state, the boundary between the fragment 49 of the substrate G and the mounting stage 13 is transformed into white; on the other hand, the shadow 50 and the occlusion mesh 39 of the substrate G are transformed into black. That is, the shadow 50 and the occlusion mesh 39 of the substrate G are not transformed into white. Therefore, it is possible to prevent the shadow 50 of the substrate G and the shielding net 39 from being changed to white and thus being judged as cracks in the substrate G, thereby enabling accurate detection of abnormalities such as cracks in the substrate G.

[0075] Furthermore, in this embodiment, as described above, the occlusion mesh 39 of the grayscale image in the pin-up state is not transformed to white. Therefore, in the binarized image, there will be no situation where it is impossible to distinguish between the small fragments 49 transformed to white and the occlusion mesh 39 transformed to white. As a result, it is possible to detect the small fragments 49 remaining on the stage 13.

[0076] Furthermore, in this embodiment, since the abnormality of the substrate G is detected based on the moving image, the abnormality of the substrate G can be detected in a relatively short time, for example, about 1 second. That is, the abnormality of the substrate G can be detected almost immediately after the pin of the substrate G is captured in the raised state. As a result, the abnormality of the substrate G can be detected after the pin of the substrate G is captured in the raised state and before the substrate G is transported to the conveying device, preventing the fragments 49 of the substrate G from still being scattered near the gate valve 20.

[0077] Furthermore, in this embodiment, the capturing of moving images is performed using plasma emission emitted by the plasma generated in the processing space U. Therefore, it is possible to capture moving images for detecting abnormalities in the substrate G without setting up a new light source for capturing images. Moreover, in this embodiment, the information processing device 40, which serves as the substrate abnormality detection device, is separate from the film deposition apparatus 10. Therefore, it can also be used for detecting the generation of abnormalities in substrates in other substrate processing apparatuses having the same structure as the film deposition apparatus 10.

[0078] Furthermore, in adaptive binarization processing, whether the central pixel is transformed into white depends on the distribution of pixel values ​​of the multiple pixels surrounding that central pixel. Therefore, the outlines of the lifting pin 38 and the fragment 49 of the substrate G may not be transformed into white without errors. For example, when a portion B of the binarized image of the evaluation object range after adaptive binarization processing is used... Figure 12 When the outline of the fragment 49 of the substrate G that was transformed into white at (A) is magnified, it can be confirmed that the outline is not smooth, and the outline has defects and protruding noise. Figure 12 (B) Therefore, if only the grayscale image of the pin in the rising state is subjected to adaptive binarization processing, the contours of the lifting pin 38 and the fragment 49 of the substrate G cannot be accurately extracted. As a result, the accuracy of anomaly detection of the substrate G using the threshold for anomaly detection may be reduced.

[0079] Therefore, it is preferable to perform morphological processing on the binarized image of the evaluation object range after adaptive binarization processing before performing anomaly detection on the substrate G using an anomaly detection threshold. Specifically, a closing operation is performed on the binarized image to reshape the contour by filling in contour defects, and then an opening operation is performed on the binarized image after the closing operation to remove protruding noise from the contour. In this embodiment, the closing operation and the opening operation are each performed at least once in the morphological processing. Alternatively, the closing operation may be performed after the opening operation in the morphological processing.

[0080] Furthermore, in the closing operation, the binarized image of the evaluation object area is subjected to erosion after dilation. Conversely, in the opening operation, the binarized image is subjected to dilation after erosion. In the dilation process, for example, as... Figure 13 As shown in (A), multiple pixels are surrounded by a structuring element (kernel) 53 in such a way that the pixel 52, which is the object of the pixel value transformation, is included. If any pixel in the structuring element 53 has a pixel value of 255 (white), then the pixel value of the pixel 52, which is the object of the pixel value transformation, is transformed to 255 (white). On the other hand, in the erosion process, for example, as... Figure 13 As shown in (B), multiple pixels are surrounded by a structuring element 55 in such a way that the pixel 54, which is the object of the pixel value transformation, is included. If there is a pixel with a pixel value of 0 (black) in the structuring element 55, the pixel value of the pixel 54, which is the object of the pixel value transformation, is transformed to 0 (black).

[0081] When morphological processing is performed on a binarized image of the evaluation object range after adaptive binarization, the outline of the fragment 49 of the substrate G after being transformed to white can be smoothed. Figure 12(C)). Then, anomaly detection of the substrate G using an anomaly detection threshold is performed (S95).

[0082] The preferred embodiments of this disclosure have been described above, but this disclosure is not limited to the above embodiments and various modifications and alterations can be made within its scope.

[0083] In the above-described embodiment, whether an anomaly in the substrate G has occurred is determined based on whether the proportion of white pixels in the binarized image of the evaluation target area is above a predetermined anomaly detection threshold. Furthermore, since the outline of the substrate G fragment 49 is considered to be composed of multiple consecutive white pixels, even if multiple white pixels are detected in the binarized image of the evaluation target area, if these multiple white pixels are not consecutive, it is considered not to be the outline of the substrate G fragment 49, and therefore no anomaly in the substrate G is considered to have occurred. Therefore, it is also possible to determine whether an anomaly in the substrate G has occurred based not only on the proportion of white pixels in the binarized image of the evaluation target area, but also on whether multiple white pixels are consecutive.

[0084] Furthermore, in this embodiment, each camera 22 captures motion images of the interior of the processing container 11 in color, but each camera 22 may also capture motion images of the interior of the processing container 11 in black and white. In this case, a grayscale image of the pin-up state can be directly obtained from the captured motion image without grayscale conversion, thereby reducing the processing load required to acquire the grayscale image.

[0085] Furthermore, in this embodiment, the substrate anomaly detection device is described as being composed of an information processing device 40 separate from the film deposition apparatus 10. However, the film deposition apparatus 10 may also possess the functions of the information processing device 40. In this case, the control unit 37 of the film deposition apparatus 10 performs... Figure 9 The substrate anomaly detection processing, in such a structure, can also achieve the same effect as in this embodiment.

[0086] Furthermore, in this embodiment, only one camera 22 for capturing moving images has been described. However, the substrate anomaly detection system 1 may also include multiple cameras 22, such as two, that capture moving images from different perspectives. In this case, if the proportion of white in either of the two binary images of the evaluation object range obtained based on the two moving images captured by the two cameras 22 is above a threshold for anomaly detection, it can be determined that the substrate G has an anomaly. Therefore, by determining whether the substrate G has an anomaly based on moving images from different perspectives, it is possible to suppress the occurrence of missed anomalies in the substrate G and improve the detection accuracy of anomalies in the substrate G.

[0087] Furthermore, in this embodiment, the camera 22 captures moving images of the interior of the processing container 11, but the camera 22 can also capture still images of the interior of the processing container 11. In this case, the camera 22 captures a still image of the interior of the processing container 11 at the moment when the pin representing the pin of the substrate G is in the raised state, which is included in the process. As a result, the process of acquiring images from moving images can be eliminated, and the processing load required to acquire grayscale images of the pin-raised state can be reduced.

[0088] In addition, in this embodiment, the film forming apparatus 10 includes a rectangular processing container 11, but the shape of the processing container 11 is not limited to a rectangular shape. For example, the processing container 11 may also be a cylindrical shape or other shapes. In this case, the substrate may also be circular.

[0089] Furthermore, this embodiment describes the application of the technology disclosed herein to a film-forming apparatus. However, the substrate processing apparatus that applies the technology disclosed herein is not limited to a film-forming apparatus; any substrate processing apparatus that performs pin lifting can apply the technology disclosed herein. Examples of such substrate processing apparatuses include etching apparatuses, ashing apparatuses, and substrate heating apparatuses. Additionally, the film-forming apparatus applying the technology disclosed herein is configured as an inductively coupled plasma processing apparatus, but it can also be configured as a capacitively coupled plasma processing apparatus, a microwave plasma processing apparatus, or other types of plasma processing apparatus.

[0090] Explanation of reference numerals in the attached figures

[0091] G: Substrate

[0092] 1: Substrate anomaly detection system

[0093] 10: Film-forming device

[0094] 11: Handling Containers

[0095] 13: Platform

[0096] 21: Monitoring Window

[0097] 22: Camera

[0098] 38: Lifting pin

[0099] 39: Screening net

[0100] 40: Information processing device

[0101] 49: Fragments

[0102] 50: Shadow

Claims

1. A substrate anomaly detection method is a method for detecting anomalies in a substrate being processed inside a processing container of a substrate processing apparatus, wherein, The substrate processing apparatus includes: A mounting stage, disposed inside the processing container, has a mounting surface for mounting the substrate; A lifting pin, which can protrude from the mounting surface and be inserted into it, can raise and lower the substrate; At least one observation window is disposed on the side wall of the processing container; as well as A control device that controls the execution of the process applied to the substrate. The substrate anomaly detection method includes the following steps: At least one camera capturing images of the interior of the processing container via the viewing window, including the state in which the substrate is detached from the mounting surface by the lifting pin, is acquired; The evaluation object range is set for the portion corresponding to the mounting surface, such that the substrate is not included in the image of the pin-raised state in which the substrate is detached from the mounting surface among the multiple images constituting the motion image. Adaptive binarization processing is applied to the evaluation object range of the grayscale image of the image representing the rising state of the pin; as well as If the proportion of white pixels in the binarized image obtained by adaptive binarization processing of the evaluation object range is above a predetermined threshold, it is determined that the substrate has an anomaly.

2. The substrate anomaly detection method according to claim 1, wherein, In the process of performing adaptive binarization processing on the evaluation object range, the outline of the substrate fragments remaining on the mounting surface included in the evaluation object range is transformed into white.

3. The substrate anomaly detection method according to claim 1, wherein, In the process of performing adaptive binarization processing on the evaluation object range, all pixels surrounded by other pixels within the evaluation object range are taken as the central pixel. For each of the central pixels, a binarization calculation region with multiple pixels surrounding the central pixel is set. For each of the binarization calculation regions, the average pixel value of each pixel in the binarization calculation region is taken as the binarization threshold. If the pixel value of the central pixel in the binarization calculation region is above the binarization threshold, the central pixel is changed to black. If the pixel value of the central pixel is less than the binarization threshold, the central pixel is changed to white.

4. The substrate anomaly detection method according to claim 3, wherein, The binarization threshold is simply the average of the pixel values ​​of the multiple pixels surrounding the central pixel in the binarization calculation region.

5. The substrate anomaly detection method according to claim 3, wherein, The binarization threshold is the average of the pixel values ​​of the plurality of pixels surrounding the central pixel in the binarization calculation region and the pixel value of the central pixel.

6. The substrate anomaly detection method according to claim 3, wherein, It also includes the following steps: adjusting the average pixel value of each pixel in the binarization calculation area according to the adjustment value to set the binarization threshold.

7. The substrate anomaly detection method according to claim 6, wherein, The adjustment value is set to a value that transforms all images of the substrate's shadow and the shielding mesh used to protect the observation window, which are included in the evaluation object range, into black during the adaptive binarization process of the evaluation object range.

8. The substrate anomaly detection method according to claim 7, wherein, In the process of setting the binarization threshold, a grayscale image of the substrate detached from the mounting surface is input into the learning-completed model, and the learning-completed model outputs the adjustment value. The learning model is generated through machine learning, which uses a combination of multiple grayscale images of a large range of evaluation objects, including the shadow of the substrate and the image of the occlusion net, and adjustment values ​​that are associated with the multiple grayscale images to transform the image into black as learning data.

9. The substrate anomaly detection method according to claim 1, wherein, It also includes the following steps: directly acquiring a grayscale image from the motion image that uses grayscale to represent the pin rising state.

10. The substrate anomaly detection method according to claim 1, wherein, The predetermined threshold varies depending on the type of substrate.

11. A substrate anomaly detection device for detecting anomalies in a substrate being processed inside a processing container of a substrate processing apparatus, wherein, The substrate processing apparatus includes: A mounting stage, disposed inside the processing container, has a mounting surface for mounting the substrate; A lifting pin, which can protrude from the mounting surface and be inserted into it, can raise and lower the substrate; At least one observation window, said at least one observation window being disposed on the side wall of the processing container; and A control device that controls the execution of the process applied to the substrate. The substrate anomaly detection device also includes a control unit. The control unit performs the following procedures: At least one camera capturing images of the interior of the processing container via the viewing window, including the state in which the substrate is detached from the mounting surface by the lifting pin, is acquired; The evaluation object range is set for the portion corresponding to the mounting surface, such that the substrate is not included in the image of the pin-raised state in which the substrate is detached from the mounting surface among the multiple images constituting the motion image. Adaptive binarization processing is applied to the evaluation object range of the grayscale image of the image representing the rising state of the pin; as well as If the proportion of white pixels in the binarized image obtained by adaptive binarization processing of the evaluation object range is above a predetermined threshold, it is determined that the substrate has an anomaly.

Images