Method for detecting the appearance of a silicon rod and related apparatus

Abstract

The application provides a silicon rod appearance detection method and related equipment, by acquiring the current first image information of the silicon rod, the first image information comprising a first image; the optimal binary threshold of the first image is obtained by the maximum inter-class variance method; the first image is binarized according to the optimal binary threshold to obtain a binary image, which can represent the current cauliflower proportion of the silicon rod; the binary image is compared with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets the standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image. In this way, the growth of the silicon rod is automatically detected, replacing manual detection of the growth of the silicon rod, and the automatic detection can set a corresponding number of detection devices according to the number of reduction furnaces, thereby achieving the beneficial effect that the silicon rod growth is detected in time when cauliflower appears.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of polysilicon production, and particularly relates to a method for detecting the appearance of a silicon rod and related equipment. BACKGROUND

[0002] In the current polysilicon production process, a variety of factors can cause the silicon rod with polysilicon attached to appear cauliflower, and a large proportion of cauliflower on the silicon rod is something that polysilicon manufacturers need to avoid.

[0003] In order to avoid a large proportion of cauliflower on the silicon rod, at present, the appearance of the silicon rod is regularly monitored by a patrol personnel using the naked eye during the production process of the silicon rod. Since a large number of reduction furnaces are usually arranged in the reduction workshop, and the silicon rods in each reduction furnace need to be monitored, this monitoring method is difficult to timely find the proportion of cauliflower on the silicon rods in each reduction furnace.

[0004] Therefore, there is a problem in the prior art that the proportion of cauliflower on the silicon rod during the growth process is difficult to be timely monitored. SUMMARY

[0005] Embodiments of the present application provide a method for detecting the appearance of a silicon rod and related equipment to solve the problem that the proportion of cauliflower on the silicon rod during the growth process is difficult to be timely monitored.

[0006] In a first aspect, the embodiments of the present application provide a method for detecting the appearance of a silicon rod, comprising:

[0007] obtaining current first image information of the silicon rod, the first image information comprising a first image;

[0008] obtaining an optimal binarization threshold of the first image by using the maximum inter-class variance method;

[0009] performing binarization processing on the first image according to the optimal binarization threshold to obtain a binary image, the binary image being capable of representing the current proportion of cauliflower on the silicon rod;

[0010] comparing the binary image with a target reference image to determine whether the current proportion of cauliflower on the silicon rod meets a standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image.

[0011] Optionally, the optimal binarization threshold of the first image is obtained by using the maximum inter-class variance method, comprising:

[0012] All pixel points of the first image are classified by a binarization threshold of the first image to obtain a plurality of category groups; wherein, pixel points with a gray value less than the binarization threshold are classified into a first category, pixel points with a gray value greater than the binarization threshold are classified into a second category, the binarization threshold of the first image has a value range of integers in 0-255, a probability of a single pixel point of all pixel points of the first image being classified into the first category is p1, a probability of the single pixel point of all pixel points of the first image being classified into the second category is p2, a mean value of pixel values of pixel points in the first category is m1, a mean value of pixel values of pixel points in the second category is m2, and a mean value of pixel values of all pixel points in the first image is mG;

[0013] According to p1, p2, m1, m2 and mG, an inter-class variance σ of the first category and the second category in each category group is calculated 2 , σ 2 satisfies: 2 = P1(m1-mG) 2 + P2(m2-mG) 2 ;

[0014] The binarization threshold corresponding to the target category group is determined as the optimal binarization threshold of the first image, and the target category group is a category group with the maximum σ 2 value in the plurality of category groups.

[0015] Optionally, before determining whether the proportion of cauliflower appearing in the silicon rod meets the standard, the method further comprises:

[0016] At least two growth images of the silicon rod are obtained, and different growth images represent different growth stages of the silicon rod;

[0017] The at least two growth images of the silicon rod are respectively subjected to binarization processing to obtain reference images of the silicon rod in different growth stages.

[0018] Optionally, the first image information further comprises growth stage information of the silicon rod when the first image is taken, and the growth stage information of the silicon rod in the binarization image is the same as the growth stage information of the silicon rod in the first image, before comparing the binarization image with the target reference image to determine whether the proportion of cauliflower appearing in the silicon rod meets the standard, the method further comprises:

[0019] Based on the growth stage information of the silicon rod when the first image is taken, a target reference image corresponding to the binarization image is determined.

[0020] Optionally, after comparing the binarization image with the target reference image to determine whether the proportion of cauliflower appearing in the silicon rod meets the standard, the method further comprises:

[0021] When it is determined that the proportion of cauliflower currently appearing on the silicon rod does not meet the standard, obtain the furnace temperature of the reduction furnace in which the silicon rod is located;

[0022] According to the furnace temperature of the reduction furnace in which the silicon rod is located and the preset furnace temperature, adjust the input current value of the reduction furnace in which the silicon rod is located or input the molar ratio of hydrogen to trichlorosilane into the reduction furnace in which the silicon rod is located.

[0023] In a second aspect, the embodiments of the present application also provide a device for detecting the appearance of a silicon rod, characterized in that the device comprises:

[0024] A first obtaining module is configured to obtain first image information of the silicon rod at present, and the first image information comprises a first image.

[0025] A second obtaining module is configured to obtain an optimal binarization threshold of the first image by using the maximum inter-class variance method.

[0026] A processing module is configured to perform binarization processing on the first image according to the optimal binarization threshold, so as to obtain a binary image, and the binary image can represent the proportion of cauliflower of the silicon rod at present.

[0027] A first determining module is configured to compare the binary image with a target reference image to determine whether the proportion of cauliflower of the silicon rod at present meets the standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image.

[0028] Optionally, the second obtaining module comprises:

[0029] A classifying submodule is configured to classify all pixel points of the first image by taking the binarization threshold of the first image as a boundary, so as to obtain a plurality of class groups; wherein the class groups comprise: the pixel points with a gray value less than the binarization threshold are classified into a first class, and the pixel points with a gray value greater than the binarization threshold are classified into a second class; the binarization threshold of the first image is an integer in the range of 0-255; the probability of a single pixel point in all pixel points of the first image being classified into the first class is p1; the probability of a single pixel point in all pixel points of the first image being classified into the second class is p2; the mean pixel value of the pixel points in the first class is m1; the mean pixel value of the pixel points in the second class is m2; and the mean pixel value of all pixel points in the first image is mG.

[0030] A calculating submodule is configured to calculate the inter-class variance σ 2 of the first class and the second class in each class group according to p1, p2, m1, m2 and mG. 2 It is satisfied that σ 2 = P1(m1-mG) 2 + P2(m2-mG) 2 .

[0031] The determining sub-module is configured to determine the binarization threshold corresponding to the target class group as the optimal binarization threshold of the first image, the target class group being a class group with the maximum value in the plurality of class groups 2 The class group with the maximum value.

[0032] Optionally, the appearance detection device of the silicon rod further includes:

[0033] The third obtaining module is configured to obtain at least two growth images of the silicon rod, different growth images representing different growth stages of the silicon rod.

[0034] The second processing module is configured to perform binarization processing on the at least two growth images of the silicon rod respectively, and obtain reference images of the silicon rod in different growth stages respectively.

[0035] Optionally, the appearance detection device of the silicon rod further includes:

[0036] The second determining module is configured to determine a target reference image corresponding to the binary image based on growth stage information of the silicon rod when the first image is captured.

[0037] Optionally, the appearance detection device of the silicon rod further includes:

[0038] The fourth obtaining module is configured to obtain a furnace temperature of a reduction furnace in which the silicon rod is located when it is determined that the cauliflower proportion of the silicon rod currently present does not meet the standard.

[0039] The adjusting module is configured to adjust an input current value of the reduction furnace in which the silicon rod is located or a molar ratio of hydrogen to trichlorosilane input into the reduction furnace in which the silicon rod is located according to the furnace temperature of the reduction furnace in which the silicon rod is located and a preset furnace temperature.

[0040] In a third aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above silicon rod appearance detection method when executing the computer program.

[0041] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program is executable on a processor to implement the steps of the above silicon rod appearance detection method.

[0042] In the embodiment of the present application, the current first image information of the silicon rod is acquired, the first image information comprising a first image; the optimal binarization threshold of the first image is acquired by the maximum inter-class variance method; the first image is binarized according to the optimal binarization threshold to obtain a binary image, the binary image representing the current cauliflower proportion of the silicon rod; the binary image is compared with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets the standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image. In this way, the growth of the silicon rod is automatically detected, replacing manual detection of the growth of the silicon rod, and the automatic detection can set a corresponding number of detection devices according to the number of reduction furnaces, thereby achieving the beneficial effect that the silicon rod is detected in time when cauliflower appears. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the description of the embodiments of the present application will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0044] Figure 1 is a flowchart of a method for appearance detection of a silicon rod provided by the embodiments of the present application;

[0045] Figure 2 is a structural schematic diagram of an appearance detection device for a silicon rod provided by the embodiments of the present application;

[0046] Figure 3 is a structural schematic diagram of an electronic device also provided by the embodiments of the present application. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0048] Unless otherwise defined, the technical terms or scientific terms used in the present application should be understood as the usual meanings understood by those skilled in the art in the field to which the present application belongs. The terms "first", "second" and similar words used in the present application do not represent any order, number or importance, but are only used to distinguish different components. "Up", "down", "left", "right" and the like are only used to represent relative positional relationships, which change accordingly when the absolute positions of the described objects change.

[0049] As Figure 1 shown in the figure, the embodiment of the present application provides a flow chart of a silicon rod appearance detection method, comprising:

[0050] Step 101, acquiring current first image information of the silicon rod, the first image information comprising a first image;

[0051] Optionally, the first image can be taken by an upper sight glass on the reduction furnace, or can be taken by a sight glass on other parts of the reduction furnace.

[0052] Optionally, the first image can be a color image or a grayscale image.

[0053] For example, the frequency of acquiring the first image information can be determined according to the growth stage of the silicon rod, for example, the growth stage of the silicon rod can be divided into early growth stage, middle growth stage and late growth stage, and the time length of each growth stage is one hour, so the frequency of acquiring the first image information is one hour each time. No further limitation is made here.

[0054] It should be noted that a plurality of silicon rods are arranged in the reaction furnace, but when the first image information is acquired, it is acquired through the sight glass, so only two silicon rods can be displayed in the picture of the first image.

[0055] Step 102, acquiring an optimal binarization threshold of the first image by the maximum inter-class variance method;

[0056] It should be noted that the optimal binarization threshold of the first image refers to that a binary image is obtained by binarization processing of the first image, and a suitable binarization threshold is required in the binarization processing process.

[0057] Step 103, binarization processing the first image according to the optimal binarization threshold to obtain a binary image, the binary image being capable of representing the current cauliflower proportion of the silicon rod;

[0058] It should be noted that the color displayed on the binary image obtained by binarization processing of the area where the cauliflower appears on the silicon rod in the first image is different from the area where no cauliflower appears on the silicon rod, so the binary image can represent the cauliflower proportion appearing on the silicon rod.

[0059] Step 104, comparing the binary image with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets the standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image.

[0060] It should be understood that the target reference image can also be a binary image, and is a reference image corresponding to the growth stage of the silicon rod in the current binary image.

[0061] Optionally, in some embodiments, the area size of the connected region can be obtained by performing binary image connected region processing analysis, and the area size of the connected region can be obtained by performing connected region analysis on the target reference image; then the area sizes of the connected regions are compared, so that it can be accurately determined whether the current cauliflower proportion meets the standard.

[0062] In the embodiment of the present application, the current first image information of the silicon rod is obtained, the first image information including a first image; the optimal binarization threshold of the first image is obtained by the maximum inter-class variance method; the first image is binarized according to the optimal binarization threshold to obtain a binary image, which can represent the current cauliflower proportion of the silicon rod; the binary image is compared with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets the standard, wherein the target reference image is a reference image corresponding to the growth stage of the silicon rod in the binary image. In this way, the growth of the silicon rod is automatically detected, replacing manual detection of the growth of the silicon rod. Automatic detection can set a corresponding number of detection devices according to the number of reduction furnaces, so as to achieve the beneficial effect that the silicon rod is detected in time when cauliflower appears.

[0063] Optionally, in some embodiments, the optimal binarization threshold of the first image is obtained by the maximum inter-class variance method, including:

[0064] All pixel points of the first image are classified with the binarization threshold of the first image as a boundary to obtain a plurality of class groups; wherein the pixel points with a gray value less than the binarization threshold are classified into a first class, the pixel points with a gray value greater than the binarization threshold are classified into a second class, the binarization threshold of the first image has a value range of integers in 0-255, the probability of a single pixel point of all pixel points of the first image being classified into the first class is p1, the probability of a single pixel point of all pixel points of the first image being classified into the second class is p2, the mean pixel value of the pixel points in the first class is m1, the mean pixel value of the pixel points in the second class is m2, and the mean pixel value of all pixel points in the first image is mG.

[0065] According to p1, p2, m1, m2 and mG, the inter-class variance σ of the first class and the second class in each class group is calculated 2 , σ 2 satisfies: 2 = P1(m1-mG) 2 + P2(m2-mG) 2 ;

[0066] The binarization threshold corresponding to the target class group is determined as the optimal binarization threshold of the first image, and the target class group is the class group with the maximum σ 2 value in the plurality of class groups.

[0067] Optionally, the value of the binarization threshold can traverse each integer in 0-255. The value of the inter-class variance of the binary image corresponding to each binarization threshold is obtained, and the binarization threshold corresponding to the maximum inter-class variance is determined as the optimal binarization threshold of the first image.

[0068] In the embodiment of the present application, the optimal binarization threshold of the first image can be obtained through the above scheme, so that the binary image (obtained after the first image is binarized) can most accurately reflect the proportion of the cauliflower on the silicon rod.

[0069] Optionally, in some embodiments, before determining whether the proportion of the cauliflower appearing during the growth of the silicon rod meets the standard by comparing the binary image with the target reference image, the method further comprises:

[0070] At least two growth images of the silicon rod are obtained, and different growth images represent different growth stages of the silicon rod.

[0071] The at least two growth images of the silicon rod are respectively binarized to obtain reference images of the silicon rod in different growth stages.

[0072] For example, three growth images of the silicon rod are obtained, which are respectively a growth image in the early growth stage, a growth image in the middle growth stage, and a growth image in the late growth stage. Then, the three growth images are respectively binarized to obtain a binary image in the early growth stage, a binary image in the middle growth stage, and a binary image in the late growth stage. Finally, the binary images are saved as reference images.

[0073] In the embodiment of the present application, the reference images can be obtained through the above scheme, which is beneficial to providing a reference for subsequent comparison of the binary image of the current silicon rod.

[0074] Optionally, in some embodiments, the first image information further comprises growth stage information of the silicon rod when the first image is taken, and the growth stage information of the silicon rod in the binary image is the same as that in the first image. Before determining whether the proportion of the cauliflower appearing in the current silicon rod meets the standard by comparing the binary image with the target reference image, the method further comprises:

[0075] Based on the growth stage information of the silicon rod when the first image is taken, the target reference image corresponding to the binary image is determined.

[0076] For example, when the first image is taken, the silicon rod is in the early growth stage. Therefore, the growth stage of the silicon rod in the binary image obtained by binarizing the first image is also in the early growth stage. When the silicon rod in the binary image is in the early growth stage, the target reference image corresponding to the binary image is the reference image of the silicon rod in the early growth stage.

[0077] Alternatively, the silicon rod may be in the middle of growth or other growth stages when the first image is taken.

[0078] For example, the reason for obtaining the growth stage information of the first image is that the standard for whether the cauliflower ratio meets the requirements varies depending on the growth stage of the silicon rod in the first image. Further, for example, the cauliflower ratio of the silicon rod in the early stage of growth cannot exceed 35%, the cauliflower ratio of the silicon rod in the middle stage of growth cannot exceed 45%, and the cauliflower ratio of the silicon rod in the later stage of growth cannot exceed 65%.

[0079] Optionally, in some embodiments, after comparing the binary image with the target reference image to determine whether the current cauliflower ratio on the silicon rod meets the standard, the method further includes:

[0080] If it is determined that the proportion of cauliflower-like growths on the silicon rod is not up to standard, the furnace temperature of the reduction furnace where the silicon rod is located is obtained.

[0081] Adjust the input current value of the reduction furnace or the molar ratio of hydrogen to trichlorosilane in the reduction furnace based on the furnace temperature and preset furnace temperature of the silicon rod.

[0082] It should be understood that when the furnace temperature is higher than the preset value, the input current value of the reduction furnace in which the silicon rod is located will be adjusted. During the growth of the silicon rod, the current value input to the reduction furnace is constantly increasing, but if the furnace temperature is higher than the preset value, the current value can be stopped from increasing, or the magnitude of the current increase can be reduced.

[0083] It should be understood that when the furnace temperature is lower than the preset value, the molar ratio of hydrogen to trichlorosilane in the reduction furnace where the silicon rod is located is increased; for example, the molar ratio of hydrogen to trichlorosilane can be increased by 0.1 to 0.5.

[0084] like Figure 2 As shown in the figure, this application embodiment also provides a structural schematic diagram of a silicon rod appearance inspection device, characterized in that the device includes:

[0085] The first acquisition module 201 is used to acquire the current first image information of the silicon rod, the first image information including the first image;

[0086] The second acquisition module 202 is used to acquire the optimal binarization threshold of the first image by means of the maximum inter-class variance method;

[0087] The processing module 203 is used to perform binarization processing on the first image according to the optimal binarization threshold to obtain a binary image, which can represent the current cauliflower ratio of the silicon rod.

[0088] The first determining module 204 is used to compare the binary image with the target reference image to determine whether the current cauliflower ratio of the silicon rod meets the standard, wherein the target reference image is the reference image corresponding to the growth stage of the silicon rod in the binary image.

[0089] Optionally, in some embodiments, the second acquisition module 202 includes:

[0090] The classification submodule is used to classify all pixels of the first image based on the binarization threshold of the first image, obtaining multiple category groups. Each category group includes pixels with gray values ​​less than the binarization threshold classified as the first category, and pixels with gray values ​​greater than the binarization threshold classified as the second category. The binarization threshold of the first image is an integer ranging from 0 to 255. The probability of a single pixel in the first image being classified into the first category is p1, and the probability of a single pixel in the first image being classified into the second category is p2. The average pixel value of pixels in the first category is m1, the average pixel value of pixels in the second category is m2, and the average pixel value of all pixels in the first image is mG.

[0091] The calculation submodule is used to calculate the inter-class variance σ of the first and second classes in each class group based on p1, p2, m1, m2, and mG. 2 , σ 2 Satisfy: σ 2 =P1(m1-mG) 2 +P2(m2-mG) 2 ;

[0092] The determination submodule is used to determine the optimal binarization threshold for the first image based on the binarization threshold corresponding to the target category group. The target category group is σ from multiple category groups. 2 The category group with the largest value.

[0093] Optionally, in some embodiments, the silicon rod appearance inspection device 200 further includes:

[0094] The third acquisition module is used to acquire at least two growth images of the silicon rod, with different growth images indicating that the silicon rod is in different growth stages;

[0095] The second processing module is used to binarize at least two growth images of the silicon rod to obtain reference images of the silicon rod at different growth stages.

[0096] Optionally, in some embodiments, the silicon rod appearance inspection device 200 further includes:

[0097] The second determining module is used to determine the target reference image corresponding to the binary image based on the growth stage information of the silicon rod when the first image was captured.

[0098] Optionally, the appearance detection device of the silicon rod further comprises:

[0099] The fourth obtaining module is configured to obtain the furnace temperature of the reduction furnace in which the silicon rod is located when it is determined that the cauliflower proportion of the current appearance of the silicon rod is not up to standard.

[0100] The adjusting module is configured to adjust the input current value of the reduction furnace in which the silicon rod is located or the molar ratio of hydrogen to trichlorosilane input into the reduction furnace in which the silicon rod is located according to the furnace temperature of the reduction furnace in which the silicon rod is located and the preset furnace temperature.

[0101] The appearance detection device 200 of the silicon rod provided by the embodiments of the present application can implement each process in the above method embodiments, and thus will not be described here again to avoid repetition.

[0102] As shown in Figure 3 The embodiments of the present application also provide a structural schematic diagram of an electronic device, which comprises a memory 302, a processor 301, and a program 3021 stored in the memory 302 and capable of running on the processor 301, and the processor 301 implements the steps of the above appearance detection method of the silicon rod when executing the program 3021.

[0103] The program 3021 can implement Figure 1 any step in the corresponding method embodiments and achieve the same beneficial effects, and thus will not be described here again.

[0104] Those skilled in the art can understand that all or part of the steps of the above method embodiments can be completed by program instructions related to hardware, and the program can be stored in a readable medium.

[0105] The embodiments of the present application also provide a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the above appearance detection method of the silicon rod.

[0106] The computer readable storage medium of an embodiment of the present application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the above. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0107] The computer readable signal medium can include a computer readable program code in a baseband or propagated as a carrier wave in a propagation medium. Such a propagated signal can take a wide variety of forms, including but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium can be any computer readable medium that can be used to carry or store computer readable program code for use by or in connection with an instruction execution system, apparatus, or device.

[0108] Program code embodied on a computer readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wire line, optical fiber cable, RF, etc., or any suitable combination of the above.

[0109] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In an embodiment of the application, the remote computer can be a server or another desktop computer.

[0110] The above describes preferred embodiments of the present application. It should be noted that, for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should also be considered as the protection scope of the present application.

Claims

1. A method of appearance inspection of a silicon rod, characterized by, The method comprises: obtaining current first image information of the silicon rod, the first image information comprising a first image; The optimal binarization threshold of the first image is obtained by the maximum inter-class variance method, and the optimal binarization threshold of the first image obtained by the maximum inter-class variance method comprises: classifying all pixel points of the first image with the binarization threshold of the first image as a boundary to obtain a plurality of class groups; wherein the class groups comprise: pixel points with a gray value less than the binarization threshold are classified into a first class, pixel points with a gray value greater than the binarization threshold are classified into a second class, the binarization threshold of the first image has an integer value in the range of 0-255, the probability of a single pixel point in all pixel points of the first image being classified into the first class is p1, the probability of a single pixel point in all pixel points of the first image being classified into the second class is p2, the mean pixel value of the pixel points in the first class is m1, the mean pixel value of the pixel points in the second class is m2, and the mean pixel value of all pixel points in the first image is mG; according to p1, p2, m1, m2 and mG, the inter-class variance σ of the first class and the second class in each class group is calculated 2 , the σ 2 satisfies: 2 σ 2 = P1(m1-mG) 2 + P2(m2-mG) 2 ; the binarization threshold corresponding to the target class group is determined as the optimal binarization threshold of the first image, and the target class group is the class group with the maximum σ in the plurality of class groups performing binarization processing on the first image according to the optimal binarization threshold to obtain a binary image, the binary image representing a current cauliflower proportion of the silicon rod; comparing the binary image with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets a standard, wherein the target reference image is a reference image corresponding to a growth stage of the silicon rod in the binary image.

2. The method of appearance inspection of a silicon ingot according to claim 1, wherein Before the comparing the binary image with the target reference image to determine whether the cauliflower proportion of the silicon rod during growth meets the standard, the method further comprises: obtaining at least two growth images of the silicon rod, different growth images representing different growth stages of the silicon rod; performing binarization processing on the at least two growth images of the silicon rod respectively to obtain reference images of the silicon rod in different growth stages respectively.

3. The method of appearance inspection of a silicon ingot according to claim 1, wherein The first image information further comprises growth stage information of the silicon rod when the first image is taken, the growth stage information of the silicon rod in the binary image being the same as the growth stage information of the silicon rod in the first image, and before the comparing the binary image with the target reference image to determine whether the current cauliflower proportion of the silicon rod meets the standard, the method further comprises: determining the target reference image corresponding to the binary image based on the growth stage information of the silicon rod when the first image is taken.

4. The silicon ingot appearance inspection method according to claim 1, wherein After the comparing the binary image with the target reference image to determine whether the cauliflower proportion of the silicon rod during growth meets the standard, the method further comprises: when it is determined that the current cauliflower proportion of the silicon rod does not meet the standard, obtaining a furnace temperature of a reduction furnace in which the silicon rod is located; adjusting an input current value of the reduction furnace in which the silicon rod is located or inputting a molar ratio of hydrogen to trichlorosilane into the reduction furnace in which the silicon rod is located according to the furnace temperature of the reduction furnace in which the silicon rod is located and a preset furnace temperature.

5. A silicon ingot appearance inspection apparatus characterized by comprising: The device comprises: a first obtaining module configured to obtain current first image information of the silicon rod, the first image information comprising a first image; The second acquisition module is configured to acquire an optimal binarization threshold of the first image by using the maximum inter-class variance method. The second acquisition module comprises: a classification submodule configured to classify all pixel points of the first image by using a binarization threshold of the first image as a boundary to obtain a plurality of class groups; wherein the class groups comprise: pixel points with a gray value less than the binarization threshold are classified into a first class, pixel points with a gray value greater than the binarization threshold are classified into a second class, the binarization threshold of the first image has an integer value in a range of 0-255, a probability of a single pixel point of all pixel points of the first image being classified into the first class is p1, a probability of a single pixel point of all pixel points of the first image being classified into the second class is p2, a mean pixel value of pixel points in the first class is m1, a mean pixel value of pixel points in the second class is m2, and a mean pixel value of all pixel points in the first image is mG; a calculation submodule configured to calculate inter-class variances of the first class and the second class in each class group according to the p1, the p2, the m1, the m2 and the mG 2 , the σ 2 satisfies: σ 2 = P1(m1-mG) 2 + P2(m2-mG) 2 ; and a determination submodule configured to determine a binarization threshold corresponding to a target class group as the optimal binarization threshold of the first image, the target class group being a class group with the maximum σ 2 . a processing module configured to perform binarization processing on the first image according to the optimal binarization threshold to obtain a binary image, the binary image representing a current cauliflower proportion of the silicon rod; a first determining module configured to compare the binary image with a target reference image to determine whether the current cauliflower proportion of the silicon rod meets a standard, wherein the target reference image is a reference image corresponding to a growth stage of the silicon rod in the binary image.

6. The silicon ingot appearance inspection apparatus according to claim 5, wherein The device further comprises: a third obtaining module configured to obtain at least two growth images of the silicon rod, different growth images representing different growth stages of the silicon rod; a second processing module configured to perform binarization processing on the at least two growth images of the silicon rod respectively to obtain reference images of the silicon rod in different growth stages respectively.

7. The silicon ingot appearance inspection apparatus according to claim 5, wherein The first image information further comprises growth stage information of the silicon rod when the first image is taken, the growth stage information of the silicon rod in the binary image being the same as the growth stage information of the silicon rod in the first image, and the device further comprises: A second determining module is configured to determine the target reference image corresponding to the binary image based on growth stage information of the silicon rod when the first image is captured.

8. The silicon ingot appearance inspection apparatus according to claim 5, wherein The device further comprises: A fourth obtaining module is configured to obtain a furnace temperature of a reduction furnace in which the silicon rod is located when it is determined that the proportion of cauliflower currently appearing on the silicon rod does not meet the standard. An adjusting module is configured to adjust an input current value of the reduction furnace in which the silicon rod is located or a molar ratio of hydrogen to trichlorosilane input into the reduction furnace according to the furnace temperature of the reduction furnace in which the silicon rod is located and a preset furnace temperature.

9. An electronic device comprising: The memory, the processor, and a program stored in the memory and capable of running on the processor; the processor is used for reading the program in the memory to implement the steps in the appearance detection method of the silicon rod according to any one of claims 1 to 4.

10. A readable storage medium for storing a program, characterized in that, The program is executed by the processor to implement the steps in the appearance detection method of the silicon rod according to any one of claims 1 to 4.

Images