WO2023226481A1 - 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置 - Google Patents

一种基于同心椭圆弦长比的单晶硅直径检测方法及装置 Download PDF

Info

Publication number
WO2023226481A1
WO2023226481A1 PCT/CN2023/076311 CN2023076311W WO2023226481A1 WO 2023226481 A1 WO2023226481 A1 WO 2023226481A1 CN 2023076311 W CN2023076311 W CN 2023076311W WO 2023226481 A1 WO2023226481 A1 WO 2023226481A1
Authority
WO
WIPO (PCT)
Prior art keywords
diameter
single crystal
ellipse
crystal silicon
length ratio
Prior art date
Application number
PCT/CN2023/076311
Other languages
English (en)
French (fr)
Inventor
项森伟
胡易人
叶敏翔
张艺菲
汪婷
门泽华
Original Assignee
之江实验室
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 之江实验室 filed Critical 之江实验室
Publication of WO2023226481A1 publication Critical patent/WO2023226481A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20032—Median filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30148—Semiconductor; IC; Wafer
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P40/00—Technologies relating to the processing of minerals
    • Y02P40/50—Glass production, e.g. reusing waste heat during processing or shaping
    • Y02P40/57—Improving the yield, e-g- reduction of reject rates

Definitions

  • the invention relates to the technical field of Czochralski single crystal silicon diameter detection, and in particular to a single crystal silicon diameter detection method and device based on the chord length ratio of concentric ellipses.
  • the Czochralski method is the main method for producing monocrystalline silicon. 70%-80% of the world's monocrystalline silicon production is produced by the Czochralski method.
  • diameter detection is an important process step and the key to ensuring the growth of crystals with equal diameters.
  • the control system measures the diameter of the single crystal silicon rod in real time, adjusts the heating power and crystal pulling speed, so that the diameter of the crystal rod is always within a reasonable range, and ultimately produces high-quality single crystal silicon rods with smooth surfaces.
  • the melting point of silicon is about 1450°C.
  • the production process of single crystal silicon is always carried out in a high temperature and negative pressure single crystal furnace. Therefore, non-contact detection methods are usually used for diameter detection.
  • the first type is the caliper method.
  • the caliper is a mechanical displacement measurement system, which consists of three parts: an optical lens, a measurement reading mechanism and an instrument base. When using it, first install the caliper on the observation window of the single crystal furnace. The operator adjusts the optical lens and visually searches for the left and right edges of the cylindrical crystal rod. After finding and aiming at one edge, move the lens to find the other edge. The distance reading of the lens movement is the diameter of the single crystal silicon.
  • This manual measurement method is more accurate, but it cannot measure in real time, continuously, and automatically.
  • the second type is the weight calculation method.
  • a weight sensor is installed on the steel wire rope that lifts the crystal rod, and the average diameter of the crystal rod within the length is calculated by calculating the increased weight of the crystal rod within the unit length.
  • This method is not a real-time detection method, and the result obtained is the average diameter of a single crystal silicon rod within a certain length.
  • the third category is machine vision method, which is also the current mainstream technology for measuring the diameter of single crystal silicon.
  • machine vision method is also the current mainstream technology for measuring the diameter of single crystal silicon.
  • the diameter of the energy release ring represents the diameter of the crystal rod, and all visual methods are developed around measuring the diameter of the energy release ring.
  • Most single crystal silicon diameter detection systems use CCD or CMOS cameras for detection. The camera is generally installed on the observation window of the crystal pulling furnace and tilts downward to photograph the energy release ring. By calculating the diameter of the energy release ring, the diameter of the single crystal silicon rod is obtained. diameter.
  • the camera Since the camera shoots from diagonally upward, the originally circular energy release ring will be deformed into an ellipse due to the shooting angle.
  • the difference between different detection methods is mainly reflected in the detection of energy release ring size.
  • the Chinese invention patent No. CN 104990510 A directly regards the edge of the oval energy release ring as a circular edge, and selects three points on the arc edge. Calculate the center of the circle and then calculate the diameter of the crystal rod. This method has fast detection speed but insufficient detection accuracy.
  • the Chinese invention patent with the public number CN 103046128 A first uses the least squares method to fit the elliptical edge of the energy release ring, and then affine the ellipse into a circular edge based on the angle between the camera and the vertical direction, and calculate the center of the circle through the coordinates of three points on the arc. and diameter. This method requires a large amount of calculation and will introduce and amplify errors in the affine transformation. The accuracy is not high either.
  • the Chinese invention patent No. CN 102914270 A proposes a support vector machine regression method for measuring the diameter of the ingot. It derives an SVR model based on the standard equation of the ellipse, and calculates the weights and offset b in the SVR model.
  • the present invention proposes a single crystal silicon diameter detection method and device based on the chord length ratio of concentric ellipses to solve the above technical problems.
  • the purpose of the present invention is to provide a single crystal silicon diameter detection method and device based on the chord length ratio of concentric ellipses, which solves the problems in the prior art that the diameter detection technology has average accuracy, large amount of calculation, and poor real-time performance.
  • a method for detecting the diameter of single crystal silicon based on the chord length ratio of concentric ellipses including the following steps:
  • Step S1 Install the industrial camera on the observation window of the crystal pulling furnace and photograph the growth process of single crystal silicon through the insulating glass;
  • Step S2 During the melting stage, perform ellipse fitting on the incomplete ellipse generated by the reflection of the guide tube, obtain the ellipse parameters, and form a complete ellipse;
  • Step S3 Preset N groups of straight lines L i passing through the center O of the ellipse, and record the intersection point of the straight line L i and the outer ellipse of the ellipse as Q i . Combine the elliptical equation and the straight line equation to solve for the coordinates of the intersection point Q i , calculate and record the Euclidean distance length of each group of OQ i ;
  • Step S4 Before entering the shoulder turning stage, set an area of interest in the energy release ring area, and the area of interest completely covers the energy release ring area in the entire crystal pulling process;
  • Step S5 In the shoulder rotation and equal diameter stages, use the edge detection algorithm to calculate the energy release ring image of the energy release ring area of the current frame, and obtain the partial energy release ring elliptical edge of the energy release ring area;
  • Step S6 Use the traversal method to query the intersection point of the line segment OQ i and the energy release ring ellipse, marked as Pi Euclidean distance length;
  • Step S7 According to the geometric properties of the chord length ratio of the concentric ellipse and combined with the diameter of the guide tube, calculate and record the single crystal silicon diameter D i corresponding to each group (OP i , OQ i );
  • Step S8 Perform median filtering on the calculated N groups of single crystal silicon diameters D i to obtain the single crystal silicon diameter corresponding to the current frame;
  • Step S9 Jump to step S5 and start the next frame of single crystal silicon diameter calculation until the single crystal silicon completes the equal diameter process.
  • the ellipse fitting algorithm in step S2 is a Hough-based ellipse fitting algorithm, a least squares ellipse fitting algorithm or a parallel chord-based ellipse fitting algorithm.
  • edge detection algorithm in step S5 is Sobel algorithm, Roberts algorithm, bilateral filtering algorithm. method or Canny edge detection algorithm.
  • the reflection of the guide tube described in step S2 and the energy release ring described in step S5 form a pair of concentric ellipses in the image.
  • the invention also provides a Czochralski single-crystal silicon diameter detection device based on the chord length ratio of concentric ellipses, including heat-insulating glass, an industrial lens, an industrial camera, a memory and one or more processors, and the memory can store Execution code, when the one or more processors execute the executable code, is used to implement a single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses described in any of the above embodiments.
  • the present invention also provides a computer-readable storage medium on which a program is stored.
  • the program is executed by a processor, the single crystal silicon diameter detection based on the chord length ratio of concentric ellipses as described in any of the above embodiments is implemented. method.
  • the present invention provides a single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses, which makes full use of the coaxial spatial relationship between the flow guide tube and the single crystal silicon rod in the vertical direction.
  • the reflection of the guide tube and the energy release ring appear as concentric ellipses in the camera imaging.
  • the diameter of single crystal silicon is calculated from the ratio of the diameter of the bottom of the flow tube and the chord length of the concentric ellipse.
  • the present invention provides a single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses.
  • the theory is simple and the operation is convenient. There is no need to fit the elliptical edge or circular edge formed by the energy release ring during diameter detection, which greatly improves the accuracy of the diameter detection. The amount of calculation is reduced and the detection accuracy and efficiency of single crystal silicon diameter are improved.
  • Figure 1 is a schematic flow chart of a single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses according to the present invention
  • Figure 2 is a schematic diagram of the internal structure of the crystal pulling furnace, the installation method of the industrial camera, the reflection of the guide tube and the energy release ring captured by the industrial camera during the production process of Czochralski single crystal silicon provided by the embodiment of the present invention;
  • Figure 3 is a schematic diagram showing a concentric elliptical relationship between the reflection of the guide tube and the energy release ring in the image according to the embodiment of the present invention
  • Figure 4 is a schematic diagram illustrating the geometric properties of the chord length ratio of concentric ellipses provided by an embodiment of the present invention
  • Figure 5 is a schematic diagram of the device for detecting the diameter of single crystal silicon based on the chord ratio of concentric ellipses according to the present invention.
  • Step S1 Install the industrial camera on the observation window of the crystal pulling furnace and photograph the growth process of single crystal silicon through the insulating glass;
  • Step S2 In the melting stage, perform ellipse fitting on the incomplete ellipse generated by the reflection of the guide tube, obtain the ellipse parameters, and form a complete ellipse;
  • the ellipse fitting algorithm is based on Hough's ellipse fitting algorithm and least squares Ellipse fitting algorithm or parallel chord based ellipse fitting algorithm;
  • Step S3 Preset N groups of straight lines L i passing through the center O of the ellipse, and record the intersection point of the straight line L i and the outer ellipse of the ellipse as Q i . Combine the elliptical equation and the straight line equation to solve for the coordinates of the intersection point Q i , calculate and record the Euclidean distance length of each group of OQ i ;
  • Step S4 Before entering the shoulder turning stage, set an ROI (Region of Interest) for the energy release ring area.
  • the ROI (Region of Interest) area can completely cover the energy release ring area during the entire crystal pulling process;
  • Step S5 In the shoulder rotation and equal diameter stages, an edge detection algorithm is used to calculate the energy release ring image of the energy release ring area of the current frame, and a partial energy release ring elliptical edge of the energy release ring area is obtained;
  • the edge detection algorithm is the Sobel algorithm. , Roberts algorithm, bilateral filtering algorithm or Canny edge detection algorithm;
  • Step S6 Use the traversal method to query the intersection point of the line segment OQ i and the energy release ring ellipse, marked as Pi Euclidean distance length;
  • Step S8 Perform median filtering on the calculated N groups of single crystal silicon diameters D i to obtain the single crystal silicon diameter corresponding to the current frame;
  • Step S9 Jump to step S5 and start the next frame of single crystal silicon diameter calculation until the single crystal silicon completes the equal diameter process.
  • the present invention installs an industrial camera 1 on the observation window 3 of the crystal pulling furnace 2.
  • the industrial camera 1 obliquely shoots the growth process of the single crystal silicon rod 5 through the insulating glass 4.
  • the high-temperature molten silicon liquid surface 6 has good flatness and strong reflectivity, and can be regarded as a mirror surface. According to the principle of mirror imaging, the guide tube 7 suspended on the molten silicon liquid surface 6 will produce a reflection, and the guide tube reflection 8 and the energy release ring 9 will be photographed and collected by the industrial camera 1 together.
  • the present invention carefully analyzes the spatial relationship between the reflection 8 of the guide tube and the energy release ring 9 in Figure 2. Since the industrial camera 1 shoots obliquely downward at an angle ⁇ with the vertical Z-axis, the cylindrical single crystal silicon rod 5 and the conical guide tube 7 become two ellipses in the image. During the assembly and production process of Czochralski single crystal silicon, the guide tube 7 and the single crystal silicon rod 5 are strictly coaxial with respect to the vertical direction, so the guide tube reflection 8 and the energy release ring 9 form a pair of concentric ellipses. Obviously, the diameter D C of the single crystal silicon rod 5 and the elliptical projection of the energy release ring 9 have the following relationship:
  • a M and b M represent the semi-major axis and the semi-minor axis of the elliptical projection of the energy release ring 9
  • k is a coefficient, which represents the conversion magnification between the pixel size and the physical size.
  • the bottom diameter D S of the guide tube 7 has the following relationship with the elliptical image of its reflection:
  • a S and b S represent the major semi-axis and minor semi-axis of the elliptical projection of the guide tube 7 .
  • Formula (3) represents that the inner ellipse and the outer ellipse have the same concentricity
  • formula (4) represents the ratio between the diameter of the single crystal silicon rod 5 and the bottom diameter of the guide tube 7 and the major semi-axes of their corresponding elliptical projections.
  • the ratios (or minor half axis ratios) are the same.
  • the diameter of the bottom of the guide tube can be measured in advance as a known constant. Therefore, as long as the major axis ratio (or minor axis ratio) of the two ellipses in the image is detected, the diameter of the crystal rod can be quickly calculated according to formula (4).
  • the ellipse has 5 parameters (x, y, a, b, ⁇ ), which respectively represent the ellipse center coordinates (x, y), the major semi-axis a, the minor semi-axis b, and the rotation angle ⁇ relative to the horizontal direction.
  • Calculating the semi-major axis a or the semi-minor axis b involves parameter identification of the ellipse. This type of problem has the disadvantages of large amount of calculation and low real-time performance.
  • the present invention deeply explores the geometric properties of concentric ellipses.
  • intersection points P i and Q i can be easily obtained from formulas (5) and (6), and the chord lengths of OP i and OQ i are respectively expressed as:
  • Formula (9) means that if the diameter of the bottom of the guide tube D S is known, the diameter of the single crystal silicon can be quickly calculated through the chord length ratio generated by the intersection of any center line with the inner ellipse and the outer ellipse.
  • the present invention also provides an embodiment of a single crystal silicon diameter detection device based on the chord length ratio of concentric ellipses. .
  • An embodiment of the present invention provides a single crystal silicon diameter detection device based on the chord length ratio of concentric ellipses, including a memory and one or more processors.
  • the memory stores executable code.
  • the one or more processors When the executable code is executed, it is used to implement a single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses in the above embodiment.
  • the embodiment of the present invention's single crystal silicon diameter detection device based on the chord ratio of concentric ellipses can be applied to any device with data processing capabilities, and any device with data processing capabilities can be a device or device such as a computer.
  • the device embodiments may be implemented by software, or may be implemented by hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it converts non-volatile data into data through the processor of any device with data processing capabilities. The corresponding computer program instructions in the permanent memory are read into the memory and run. From the hardware level, as shown in Figure 5, it is a hardware structure diagram of any device with data processing capabilities where an Internet of Things device collaborative linkage device of the present invention is located.
  • any device with data processing capabilities where the device in the embodiment is located may also include other hardware based on the actual functions of any device with data processing capabilities, which will not be described again.
  • the device embodiment since it basically corresponds to the method embodiment, please refer to the partial description of the method embodiment for relevant details.
  • the device embodiments described above are only illustrative.
  • the units described as separate components may or may not be physically separated.
  • the components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present invention. Persons of ordinary skill in the art can understand and implement the method without any creative effort.
  • An embodiment of the present invention also provides a computer-readable storage medium on which a program is stored.
  • the program is executed by a processor, the single crystal silicon diameter detection method based on the chord length ratio of concentric ellipses in the above embodiment is implemented.
  • the computer-readable storage medium may be an internal storage unit of any device with data processing capabilities as described in any of the foregoing embodiments, such as a hard disk or a memory.
  • the computer-readable storage medium can also be an external storage device of any device with data processing capabilities, such as a plug-in hard disk, smart memory card (Smart Media Card, SMC), SD card, flash memory card equipped on the device (Flash Card) etc.
  • the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capabilities.
  • the computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • Geometry (AREA)
  • Multimedia (AREA)
  • Crystals, And After-Treatments Of Crystals (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

本发明公开了一种基于同心椭圆弦长比的单晶硅直径检测方法及装置，步骤S1：拍摄单晶硅生长过程；步骤S2：椭圆拟合，获取椭圆参数；步骤S3：联立椭圆方程和直线方程；步骤S4:对释能环区域设置ROI（感兴趣区域）；步骤S5:采用边缘检测算法计算得到释能环区域的部分椭圆边缘；步骤S6:使用遍历法查询并记录距离长度；步骤S7:根据同心椭圆弦长比的几何性质，计算并记录每一组对应的单晶硅直径；步骤S8：对于计算得到的单晶硅直径进行中值滤波，得出当前帧对应的单晶硅直径；步骤S9：跳转步骤S5，开启下一帧直至等径工艺后结束。本发明在直径检测时不需要拟合释能环形成的椭圆边缘或圆边缘，极大地减少计算量，提高单晶硅直径的检测精度和检测效率。

Description

一种基于同心椭圆弦长比的单晶硅直径检测方法及装置 技术领域
本发明涉及一种直拉式单晶硅直径检测技术领域，尤其涉及一种基于同心椭圆弦长比的单晶硅直径检测方法及装置。
背景技术
信息和能源是21世纪的两大支柱产业。作为一种良好的半导体材料，单晶硅通常用于制作集成电路和太阳能电池，是信息产业和新能源产业最基础的原材料，在军事电子设备中也发挥着重要的作用。直拉法是生产单晶硅的主要方法，世界上70％-80％的单晶硅产量是用直拉法生产的。使用直拉法生产单晶硅时，直径检测是重要的工艺环节，是保证晶体等径生长的关键。控制系统通过实时测量单晶硅棒的直径，调整加热功率和晶体拉速，使晶棒直径始终处于合理范围，最终可以生产出表面光滑的高品质单晶硅棒。
硅的熔点约为1450℃，单晶硅的生产过程始终在高温负压的单晶炉内进行，因此直径检测通常采用非接触式检测手段，常用的检测方法有三类。第一类是测径仪法，测径仪是一种机械式位移测量系统，由光学镜头、测量读数机构和仪器座三部分组成。使用时，先将测径仪安装在单晶炉观察窗口上，操作人员调整光学镜头，通过目视寻找圆柱晶棒的左右边缘。找到并瞄准一条边缘后，移动镜头寻找另一条边缘，镜头移动的距离读数就是单晶硅的直径。这种人工测量方法比较准确，但无法实时、持续、自动测量。
第二类是重量计算法，在拉升晶棒的钢丝绳上安装一个重量传感器，通过计算单位长度内晶棒增加的重量来计算该长度内晶棒的平均直径。这种方法不是一种实时的检测方法，得到的结果是单晶硅棒一定长度内的平均直径。
第三类是机器视觉法，也是目前测量单晶硅直径的主流技术。单晶硅生长过程中，晶体从液态转化为固态时会释放大量能量，从而在晶棒和熔硅液面的交界处形成一个明亮的圆环，又称释能环。释能环的直径代表了晶棒的直径，所有的视觉方法都是围绕测量释能环直径进行展开。大多数单晶硅直径检测系统都使用CCD或CMOS相机进行检测，相机一般安装在拉晶炉的观察窗口上，倾斜向下拍摄释能环，通过计算释能环直径，得到单晶硅棒的直径。由于相机是从斜上方向下拍摄，原本呈圆形的释能环会因为拍摄角度变形为椭圆。不同检测方法之间的差异主要体现在释能环尺寸检测环节，如公号CN 104990510 A的中国发明专利直接将椭圆形的释能环边缘视为圆边缘，通过选取圆弧边缘上三个点计算圆心，进而计算晶棒直径。这个方法检测速度快，但是检测精度不足。公号CN 103046128 A的中国发明专利先用最小二乘法拟合释能环的椭圆边缘，然后根据相机与垂直方向的夹角，将椭圆仿射成圆边缘，通过圆弧上三点坐标计算圆心和直径。这个方法计算量大，且在仿射变换中会引入和放大误差， 精度也不高。公号CN 102914270 A的中国发明专利提出了一种支持向量机回归的晶棒直径测量方法，它针对椭圆的标准方程推导出一个SVR模型，通过求取SVR模型中的权值和偏移量b来计算椭圆拟合的参数，进一步计算晶棒直径。这个方法检测精度较高，但是计算量大，因为确定一个椭圆需要5个参数，因此实际应用时实时性较差。
为此，本发明提出一种基于同心椭圆弦长比的单晶硅直径检测方法及装置以此解决上述技术问题。
发明内容
本发明的目的在于提供一种基于同心椭圆弦长比的单晶硅直径检测方法及装置，解决了现有技术中直径检测技术精度一般、计算量大、实时性差的问题。
本发明采用的技术方案如下：
一种基于同心椭圆弦长比的单晶硅直径检测方法，包括以下步骤：
步骤S1：将工业相机安装在拉晶炉观察窗口上，透过隔热玻璃拍摄单晶硅生长过程；
步骤S2：在融料阶段，对导流筒倒影产生的残缺椭圆进行椭圆拟合，获取椭圆参数，并形成完整的椭圆；
步骤S3：预先设置N组经过所述椭圆的圆心O的直线Li,并记直线Li与所述椭圆的外椭圆交点为Qi，联立椭圆方程和直线方程，求解交点Qi的坐标，计算并记录每一组OQi的欧式距离长度；
步骤S4：进入转肩阶段前，对释能环区域设置感兴趣区域，所述感兴趣区域完全覆盖整个拉晶过程中的释能环区域；
步骤S5：在转肩、等径阶段，采用边缘检测算法对当前帧释能环区域的释能环图像计算，得到释能环区域的部分释能环椭圆边缘；
步骤S6：使用遍历法查询线段OQi与所述释能环椭圆的交点，记为Pi,根据所述释能环椭圆的圆心O和交点Pi的坐标计算并记录每一组OPi的欧式距离长度；
步骤S7：根据同心椭圆弦长比的几何性质，结合导流筒直径，计算并记录每一组(OPi，OQi)对应的单晶硅直径Di；
步骤S8：对计算得到的N组单晶硅直径Di进行中值滤波，得出当前帧对应的单晶硅直径；
步骤S9：跳转步骤S5，开启下一帧单晶硅直径计算，直至单晶硅完成等径工艺后结束。
进一步地，所述步骤S2中椭圆拟合的算法为基于霍夫的椭圆拟合算法、最小二乘椭圆拟合算法或基于平行弦的椭圆拟合算法。
进一步地，所述步骤S5中所述边缘检测算法为Sobel算法、Roberts算法、双边滤波算 法或Canny边缘检测算法。
进一步地，步骤S2中所述的导流筒倒影、步骤S5中所述的释能环在图像中形成一对同心椭圆。
进一步地，所述步骤S7中所述的同心椭圆弦长比的几何性质，指的是单晶硅直径＝导流筒底部直径×OPi的欧式距离长度/OQi的欧式距离长度。
本发明还提供一种基于同心椭圆弦长比的直拉式单晶硅直径检测装置，包括隔热玻璃、工业镜头、工业相机、存储器和一个或多个处理器，所述存储器中存储有可执行代码，所述一个或多个处理器执行所述可执行代码时，用于实现上述实施例任一项所述的一种基于同心椭圆弦长比的单晶硅直径检测方法。
本发明还提供一种计算机可读存储介质，其上存储有程序，该程序被处理器执行时，实现上述实施例任一项所述的一种基于同心椭圆弦长比的单晶硅直径检测方法。
本发明的有益效果是：
1、与现有技术比，本发明提供的一种基于同心椭圆弦长比的单晶硅直径检测方法，充分利用了导流筒和单晶硅棒在垂直方向同轴的空间关系，深入分析得出导流筒倒影和释能环在相机成像中呈现同心椭圆，进一步探索同心椭圆中心线弦长比例关系，推导得出单晶硅直径与导流筒底部直径的线性映射规律，进而通过导流筒底部直径和同心椭圆弦长比计算单晶硅直径。
2、本发明所提供的一种基于同心椭圆弦长比的单晶硅直径检测方法，理论简单，操作便利，在直径检测时不需要拟合释能环形成的椭圆边缘或圆边缘，极大地减少了计算量，提高了单晶硅直径的检测精度和检测效率。
附图说明
图1为本发明一种基于同心椭圆弦长比的单晶硅直径检测方法流程示意图；
图2为本发明实施例提供的直拉式单晶硅生产过程中，拉晶炉内结构、工业相机安装方式、工业相机拍摄的导流筒倒影和释能环的成像示意图；
图3为本发明实施例提供的导流筒倒影和释能环在图像中呈同心椭圆关系的示意图；
图4为本发明实施例提供的同心椭圆弦长比率几何性质的原理示意图；
图5位本发明一种基于同心椭圆弦长比的单晶硅直径检测方法装置示意图。
附图标记说明
1-工业相机，2-拉晶炉，3-观察窗，4-隔热玻璃，5-晶硅棒，6-熔硅液面，7-导流筒，8-
导流筒倒影，9-释能环。
具体实施方式
以下对至少一个示例性实施例的描述实际上仅仅是说明性的，决不作为对本发明及其应用或使用的任何限制。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
见图1，一种基于同心椭圆弦长比的单晶硅直径检测方法，包括以下步骤：
步骤S1：将工业相机安装在拉晶炉观察窗口上，透过隔热玻璃拍摄单晶硅生长过程；
步骤S2：在融料阶段，对导流筒倒影产生的残缺椭圆进行椭圆拟合，获取椭圆参数，并形成完整的椭圆；椭圆拟合的算法为基于霍夫的椭圆拟合算法、最小二乘椭圆拟合算法或基于平行弦的椭圆拟合算法；
步骤S3：预先设置N组经过所述椭圆的圆心O的直线Li，并记直线Li与所述椭圆的外椭圆交点为Qi，联立椭圆方程和直线方程，求解交点Qi的坐标，计算并记录每一组OQi的欧式距离长度；
步骤S4：进入转肩阶段前，对释能环区域设置ROI(感兴趣区域)，ROI(感兴趣区域)区域能完全覆盖整个拉晶过程中的释能环区域；
步骤S5：在转肩、等径阶段，采用边缘检测算法对当前帧释能环区域的释能环图像计算，得到释能环区域的部分释能环椭圆边缘；所述边缘检测算法为Sobel算法、Roberts算法、双边滤波算法或Canny边缘检测算法；
步骤S6：使用遍历法查询线段OQi与所述释能环椭圆的交点，记为Pi,根据所述释能环椭圆的圆心O和交点Pi的坐标计算并记录每一组OPi的欧式距离长度；
步骤S7：根据同心椭圆弦长比的几何性质，结合导流筒直径，计算并记录每一组(OPi，OQi)对应的单晶硅直径Di；所述的同心椭圆弦长比的几何性质，指的是单晶硅直径＝导流筒底部直径×(OPi的欧式距离长度/OQi的欧式距离长度)；
步骤S8：对计算得到的N组单晶硅直径Di进行中值滤波，得出当前帧对应的单晶硅直径；
步骤S9：跳转步骤S5，开启下一帧单晶硅直径计算，直至单晶硅完成等径工艺后结束。
实施例
如图2所示，本发明将一个工业相机1安装在拉晶炉2的观察窗3上，工业相机1透过隔热玻璃4斜向下拍摄单晶硅棒5的生长过程。高温的熔硅液面6具备较好的平整度和较强的反射性，可以视为镜面。根据镜面成像原理，悬挂在熔硅液面6上的导流筒7会产生一个倒影，导流筒倒影8和释能环9会被工业相机1一同拍摄采集。
如图3所示，本发明对图2中导流筒倒影8和释能环9的空间关系进行仔细分析。由于工业相机1与垂直方向Z轴呈β角度斜向下拍摄，圆柱形单晶硅棒5和圆锥形导流筒7在图像上变为两个椭圆。在直拉式单晶硅装配和生产过程中，导流筒7和单晶硅棒5关于垂直方向严格同轴，因此导流筒倒影8和释能环9构成一对同心椭圆。显然，单晶硅棒5的直径DC与释能环9的椭圆投影存在如下关系：
其中aM和bM表示释能环9的椭圆投影的长半轴和短半轴，k是系数，表示像素尺寸与物理尺寸间的转换倍率。同样，导流筒7底部直径DS与其倒影的椭圆图像存在如下关系：
其中aS和bS表示导流筒7椭圆投影的长半轴和短半轴。
结合公式(1)和公式(2)，容易推导出：

公式(3)表示内椭圆和外椭圆具有相同的同心率，公式(4)表示单晶硅棒5的直径和导流筒7的底部直径之间的比率与它们对应的椭圆投影的长半轴比率(或者短半轴比率)相同。导流筒底部直径可以事先测量作为已知常量，因此只要检测出图像中两个椭圆的长轴比(或短轴比)，就能根据公式(4)快速计算晶棒的直径。
椭圆有5个参数(x,y,a,b,θ)，分别表示椭圆圆心坐标(x,y)、长半轴a、短半轴b、相对水平方向的旋转角θ。想要计算长半轴a或短半轴b涉及到椭圆的参数辨识，该类问题存在计算量大、实时性低的缺点。为了减少直径检测的计算量，进一步提高检测效率，本发明对同心椭圆的几何性质进行了深入挖掘。
如图4所示是一对未经旋转、圆心在原点，具有典型代表意义的同心椭圆。椭圆A的五个参数分别是(0,0,aA,bA,00)，椭圆B的五个参数为(0,0,aB,bB,00)。经过圆心以任意倾斜角度αi画一条中心线li，交椭圆A和B分别于点Pi和Qi.其中，椭圆A和椭圆B的方程可以分别表示为：
中心线li的方程表示为：
y＝tan(αi)×x          (6)
由公式(5)和(6)易得交点Pi和Qi的位置，进而OPi和OQi的弦长分别表示为：
如公式(3)所示，同心椭圆拥有相同的同心率，那么OPi和OQi的弦长比表示为：
结合公式(4)和(8)可得：
公式(9)意味着，已知导流筒底部直径DS，可以通过任意一条中心线与内椭圆和外椭圆相交产生的弦长比值，快速计算单晶硅直径。
参见图5，与前述一种基于同心椭圆弦长比的单晶硅直径检测方法的实施例相对应，本发明还提供了一种基于同心椭圆弦长比的单晶硅直径检测装置的实施例。
本发明实施例提供的一种基于同心椭圆弦长比的单晶硅直径检测装置，包括存储器和一个或多个处理器，所述存储器中存储有可执行代码，所述一个或多个处理器执行所述可执行代码时，用于实现上述实施例中的一种基于同心椭圆弦长比的单晶硅直径检测方法。
本发明一种基于同心椭圆弦长比的单晶硅直径检测装置的实施例可以应用在任意具备数据处理能力的设备上，该任意具备数据处理能力的设备可以为诸如计算机等设备或装置。装置实施例可以通过软件实现，也可以通过硬件或者软硬件结合的方式实现。以软件实现为例，作为一个逻辑意义上的装置，是通过其所在任意具备数据处理能力的设备的处理器将非易失 性存储器中对应的计算机程序指令读取到内存中运行形成的。从硬件层面而言，如图5所示，为本发明一种物联网设备协同联动装置所在任意具备数据处理能力的设备的一种硬件结构图，除了图5所示的处理器、内存、网络接口、以及非易失性存储器之外，实施例中装置所在的任意具备数据处理能力的设备通常根据该任意具备数据处理能力的设备的实际功能，还可以包括其他硬件，对此不再赘述。
上述装置中各个单元的功能和作用的实现过程具体详见上述方法中对应步骤的实现过程，在此不再赘述。
对于装置实施例而言，由于其基本对应于方法实施例，所以相关之处参见方法实施例的部分说明即可。以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本发明方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
本发明实施例还提供一种计算机可读存储介质，其上存储有程序，该程序被处理器执行时，实现上述实施例中的一种基于同心椭圆弦长比的单晶硅直径检测方法。
所述计算机可读存储介质可以是前述任一实施例所述的任意具备数据处理能力的设备的内部存储单元，例如硬盘或内存。所述计算机可读存储介质也可以是任意具备数据处理能力的设备的外部存储设备，例如所述设备上配备的插接式硬盘、智能存储卡(Smart Media Card，SMC)、SD卡、闪存卡(Flash Card)等。进一步的，所述计算机可读存储介质还可以既包括任意具备数据处理能力的设备的内部存储单元也包括外部存储设备。所述计算机可读存储介质用于存储所述计算机程序以及所述任意具备数据处理能力的设备所需的其他程序和数据，还可以用于暂时地存储已经输出或者将要输出的数据。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (7)

  1. 一种基于同心椭圆弦长比的单晶硅直径检测方法，其特征在于，包括以下步骤：
    步骤S1：将工业相机安装在拉晶炉观察窗口上，透过隔热玻璃拍摄单晶硅生长过程；
    步骤S2：在融料阶段，对导流筒倒影产生的残缺椭圆进行椭圆拟合，获取椭圆参数，并形成完整的椭圆；
    步骤S3：预先设置N组经过所述椭圆的圆心O的直线Li，并记直线Li与所述椭圆的外椭圆交点为Qi，联立椭圆方程和直线方程，求解交点Qi的坐标，计算并记录每一组OQi的欧式距离长度；
    步骤S4：进入转肩阶段前，对释能环区域设置感兴趣区域，所述感兴趣区域完全覆盖整个拉晶过程中的释能环区域；
    步骤S5：在转肩、等径阶段，采用边缘检测算法对当前帧释能环区域的释能环图像计算，得到释能环区域的部分释能环椭圆边缘；
    步骤S6：使用遍历法查询线段OQi与所述释能环椭圆的交点，记为Pi,根据所述释能环椭圆的圆心O和交点Pi的坐标计算并记录每一组OPi的欧式距离长度；
    步骤S7：根据同心椭圆弦长比的几何性质，结合导流筒直径，计算并记录每一组(OPi，OQi)对应的单晶硅直径Di；
    步骤S8：对计算得到的N组单晶硅直径Di进行中值滤波，得出当前帧对应的单晶硅直径；
    步骤S9：跳转步骤S5，开启下一帧单晶硅直径计算，直至单晶硅完成等径工艺后结束。
  2. 如权利要求1所述的一种基于同心椭圆弦长比的单晶硅直径检测方法，其特征在于，所述步骤S2中椭圆拟合的算法为基于霍夫的椭圆拟合算法、最小二乘椭圆拟合算法或基于平行弦的椭圆拟合算法。
  3. 如权利要求1所述的一种基于同心椭圆弦长比的单晶硅直径检测方法，其特征在于，所述步骤S5中所述边缘检测算法为Sobel算法、Roberts算法、双边滤波算法或Canny边缘检测算法。
  4. 如权利要求1所述的一种基于同心椭圆弦长比的单晶硅直径检测方法，其特征在于，步骤S2中所述的导流筒倒影、步骤S5中所述的释能环在图像中形成一对同心椭圆。
  5. 如权利要求1所述的一种基于同心椭圆弦长比的单晶硅直径检测方法，其特征在于，所述步骤S7中所述的同心椭圆弦长比的几何性质，指的是单晶硅直径＝导流筒底部直径×(OPi的欧式距离长度/OQi的欧式距离长度)。
  6. 一种基于同心椭圆弦长比的直拉式单晶硅直径检测装置，其特征在于，包括隔热玻璃、工业镜头、工业相机、存储器和一个或多个处理器，所述存储器中存储有可执行代码，所述 一个或多个处理器执行所述可执行代码时，用于实现权利要求1-5中任一项所述的一种基于同心椭圆弦长比的单晶硅直径检测方法。
  7. 一种计算机可读存储介质，其特征在于，其上存储有程序，该程序被处理器执行时，实现权利要求1-5中任一项所述的一种基于同心椭圆弦长比的单晶硅直径检测方法。
PCT/CN2023/076311 2022-05-26 2023-02-16 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置 WO2023226481A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210577992.9A CN114926440A (zh) 2022-05-26 2022-05-26 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置
CN202210577992.9 2022-05-26

Publications (1)

Publication Number Publication Date
WO2023226481A1 true WO2023226481A1 (zh) 2023-11-30

Family

ID=82810338

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/076311 WO2023226481A1 (zh) 2022-05-26 2023-02-16 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置

Country Status (2)

Country Link
CN (1) CN114926440A (zh)
WO (1) WO2023226481A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114926440A (zh) * 2022-05-26 2022-08-19 之江实验室 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置
CN115187599B (zh) * 2022-09-09 2022-12-13 之江实验室 基于几何性质的轻量化的单晶硅椭圆参数辨识方法和系统
CN115265391B (zh) * 2022-09-30 2023-02-17 杭州利珀科技有限公司 一种单晶硅棒制备过程中硅料液距检测方法
CN117187942B (zh) * 2023-09-11 2024-03-26 保定景欣电气有限公司 一种拉晶过程中坩埚位置控制方法及装置
CN117385459A (zh) * 2023-10-23 2024-01-12 保定景欣电气有限公司 一种晶体生长过程中结晶检测方法及装置

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102061517A (zh) * 2010-12-13 2011-05-18 浙江长兴众成电子有限公司 直拉单晶硅直径测量方法
US20110255792A1 (en) * 2009-10-20 2011-10-20 Canon Kabushiki Kaisha Information processing apparatus, control method for the same, and computer-readable storage medium
CN112116667A (zh) * 2020-09-22 2020-12-22 扬州大学 一种发动机表面加工孔直径测量算法
CN112381807A (zh) * 2020-11-18 2021-02-19 北京图知天下科技有限责任公司 一种直拉单晶生产中晶体直径检测方法、系统及计算机
US20210196101A1 (en) * 2018-09-21 2021-07-01 Fujifilm Corporation Image processing apparatus and image processing method
CN113888572A (zh) * 2021-09-22 2022-01-04 西北工业大学 一种视觉的平面孔测量方法
CN114370828A (zh) * 2021-12-28 2022-04-19 中国铁路设计集团有限公司 基于激光扫描的盾构隧道直径收敛和径向错台检测方法
CN114399489A (zh) * 2022-01-12 2022-04-26 苏州天准科技股份有限公司 拉晶过程中光圈直径的监测方法、存储介质和终端
CN114926440A (zh) * 2022-05-26 2022-08-19 之江实验室 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110255792A1 (en) * 2009-10-20 2011-10-20 Canon Kabushiki Kaisha Information processing apparatus, control method for the same, and computer-readable storage medium
CN102061517A (zh) * 2010-12-13 2011-05-18 浙江长兴众成电子有限公司 直拉单晶硅直径测量方法
US20210196101A1 (en) * 2018-09-21 2021-07-01 Fujifilm Corporation Image processing apparatus and image processing method
CN112116667A (zh) * 2020-09-22 2020-12-22 扬州大学 一种发动机表面加工孔直径测量算法
CN112381807A (zh) * 2020-11-18 2021-02-19 北京图知天下科技有限责任公司 一种直拉单晶生产中晶体直径检测方法、系统及计算机
CN113888572A (zh) * 2021-09-22 2022-01-04 西北工业大学 一种视觉的平面孔测量方法
CN114370828A (zh) * 2021-12-28 2022-04-19 中国铁路设计集团有限公司 基于激光扫描的盾构隧道直径收敛和径向错台检测方法
CN114399489A (zh) * 2022-01-12 2022-04-26 苏州天准科技股份有限公司 拉晶过程中光圈直径的监测方法、存储介质和终端
CN114926440A (zh) * 2022-05-26 2022-08-19 之江实验室 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置

Also Published As

Publication number Publication date
CN114926440A (zh) 2022-08-19

Similar Documents

Publication Publication Date Title
WO2023226481A1 (zh) 一种基于同心椭圆弦长比的单晶硅直径检测方法及装置
CN107576281B (zh) 一种测量管路弯曲半径的方法和装置
WO2021012445A1 (zh) 一种液口距确定方法、装置及单晶炉
CN107680156B (zh) 基于偏振信息的三维重建方法
CN106157246B (zh) 一种全自动的快速柱面全景图像拼接方法
CN108828554B (zh) 无需激光落点的基于坐标变换的测量方法、系统及装置
CN107492123B (zh) 一种利用路面信息的道路监控摄像机自标定方法
CN103902953B (zh) 一种屏幕检测系统及方法
CN109579695B (zh) 一种基于异构立体视觉的零件测量方法
CN106871787A (zh) 大空间线扫描成像三维测量方法
CN110223355B (zh) 一种基于双重极线约束的特征标志点匹配方法
CN110879080A (zh) 一种高温锻件高精度智能测量仪和测量方法
CN110146030A (zh) 基于棋盘格标志法的边坡表面变形监测系统和方法
CN107862713B (zh) 针对轮询会场的摄像机偏转实时检测预警方法及模块
CN102221331A (zh) 一种基于不对称双目立体视觉技术的测量方法
WO2021115248A1 (zh) 酵母分析方法
CN107121070A (zh) 一种胶阀针头标定系统及标定方法
WO2019232793A1 (zh) 双摄像头标定方法、电子设备、计算机可读存储介质
CN103308000A (zh) 基于双目视觉的曲线物体测量方法
WO2019127319A1 (zh) 头戴式显示设备的畸变测量方法及系统
WO2021035882A1 (zh) 一种采用鱼眼镜头的声源定位方法及其设备
Wang et al. A novel method for diameter measurement of silicon single crystal
CN112508885B (zh) 一种弯管的三维中轴线检测方法及系统
CN105758337A (zh) 一种获取透镜平面与图像传感器平面之间夹角的方法
CN108353187B (zh) 用于对由光学采集系统捕获的图像进行编码的装置和方法

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2023542843

Country of ref document: JP

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23810557

Country of ref document: EP

Kind code of ref document: A1