WO2021208324A1 - 一种基于图像的测温取样下枪位置智能检测方法及系统 - Google Patents

一种基于图像的测温取样下枪位置智能检测方法及系统 Download PDF

Info

Publication number
WO2021208324A1
WO2021208324A1 PCT/CN2020/112190 CN2020112190W WO2021208324A1 WO 2021208324 A1 WO2021208324 A1 WO 2021208324A1 CN 2020112190 W CN2020112190 W CN 2020112190W WO 2021208324 A1 WO2021208324 A1 WO 2021208324A1
Authority
WO
WIPO (PCT)
Prior art keywords
gun
temperature measurement
image
sampling
steel slag
Prior art date
Application number
PCT/CN2020/112190
Other languages
English (en)
French (fr)
Inventor
龚贵波
刘向东
赖俊霖
刘贵林
刘景亚
Original Assignee
中冶赛迪工程技术股份有限公司
中冶赛迪技术研究中心有限公司
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 中冶赛迪工程技术股份有限公司, 中冶赛迪技术研究中心有限公司 filed Critical 中冶赛迪工程技术股份有限公司
Priority to DE112020000297.7T priority Critical patent/DE112020000297T5/de
Publication of WO2021208324A1 publication Critical patent/WO2021208324A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
    • G01K13/00—Thermometers specially adapted for specific purposes
    • G01K13/12—Thermometers specially adapted for specific purposes combined with sampling devices for measuring temperatures of samples of materials
    • G01K13/125—Thermometers specially adapted for specific purposes combined with sampling devices for measuring temperatures of samples of materials for siderurgical purposes
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J11/00—Manipulators not otherwise provided for
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1602—Programme controls characterised by the control system, structure, architecture
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
    • B25J9/1697—Vision controlled systems

Definitions

  • the invention relates to the field of metallurgy, in particular to an image-based method and system for intelligent detection of gun position under temperature measurement and sampling.
  • the manipulator usually uses a fixed path to complete the operation.
  • the gun head may damage the gun head due to the collision with the steel slag, resulting in failure to meet the requirements of temperature measurement and sampling.
  • a method is needed. The new detection method detects the distribution of steel slag in the area of the lower gun point, and accurately locates the steel slag gap area suitable for the lower gun.
  • the present invention provides an image-based method and system for intelligent detection of gun position under temperature measurement sampling to solve the above technical problems.
  • the intelligent detection method of gun position based on image-based temperature measurement and sampling includes:
  • the target image being an image of a molten steel surface
  • the preset temperature measurement sampling gun position, the area of interest around the gun position, and the smallest area of the gun are adapted to the obtained steel slag gap width, and the gun position that meets the conditions of the gun is selected from the molten steel level Take temperature sampling.
  • the contour information of the steel slag crack is acquired according to a preset grayscale threshold.
  • the gray threshold includes a low tail value and a high tail value
  • the area is a steel slag area
  • the area is a molten steel area
  • the area is a mixed zone of molten steel with both slag and molten steel.
  • an edge search is performed on the steel slag gap area, the inner contour and the outer contour of the steel slag gap area are obtained, and the distance between the inner contour and the outer contour is used as the steel slag gap width.
  • the preset temperature measurement sampling gun position, the area of interest around the gun position and the smallest area of the gun are adapted to the obtained steel slag gap width, and the temperature measurement is performed according to the preset gun down strategy Sampling, the preset shooting strategy includes the closest point strategy and/or the most suitable point strategy, where,
  • the closest point strategy includes selecting the smallest area in the neighborhood around the preset temperature measurement sampling gun position that satisfies the gun down position as the gun down position;
  • the most suitable point strategy includes selecting the area with the largest width of the steel slag gap in the area of interest of the preset temperature measurement sampling gun position as the gun position.
  • the preset gun dropping strategy further includes that when the closest point strategy is used to find a position that meets the condition and fails, the most suitable point strategy is adopted to determine the gun dropping position, and temperature measurement and sampling are performed.
  • the present invention also provides an image-based intelligent detection system for gun position under temperature measurement and sampling, which includes
  • An image acquisition module for acquiring a target image, the target image being an image of the molten steel surface
  • An image processing module configured to obtain contour information of the steel slag crack in the molten steel surface according to the collected target image, the contour information including the inner and outer contours of the steel slag crack; and obtain the width of the steel slag crack according to the contour information of the steel slag crack;
  • the temperature measurement sampling module is used to adapt the preset temperature measurement sampling gun position, the area of interest around the gun position, and the minimum area of the gun to the obtained steel slag gap width, and select the steel liquid level to meet the requirements Take temperature measurement and sampling at the position of the gun under the gun condition.
  • an image acquisition protection module which includes a cooling unit for reducing the temperature of the working environment of the image acquisition system and a protective cover for heat insulation, radiation prevention, and dust prevention.
  • the installation module includes an installation bracket and a temperature measurement sampling manipulator, and the image acquisition module is fixed to the temperature measurement via the installation bracket The mobile end of the sampling robot.
  • the image-based temperature measurement sampling method and system for intelligent detection of the gun position in the present invention detects the distribution of steel slag on the molten steel surface of the ladle through images, and calculates all the preset gun down areas.
  • the width of the steel slag gap is selected through a predetermined strategy to select the lower gun position that meets the lower gun conditions to perform temperature measurement and sampling, thereby improving the intelligent level of the manipulator's temperature measurement and sampling operation.
  • FIG. 1 is a schematic flowchart of an image-based method for intelligent detection of gun position under temperature measurement and sampling in an embodiment of the present invention.
  • Fig. 2 is a schematic structural diagram of an intelligent detection system for gun position based on image-based temperature measurement and sampling in an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of the target detection image and the region of interest of the intelligent detection system for the gun position under the image-based temperature measurement sampling in the embodiment of the present invention.
  • Fig. 4 is a grayscale histogram of a region of interest of the gun position intelligent detection system based on image-based temperature measurement sampling in an embodiment of the present invention.
  • Fig. 5 is a binary image inner and outer contour diagram of a region of interest of a gun position intelligent detection system based on image-based temperature measurement sampling in an embodiment of the present invention.
  • FIG. 6 is a schematic diagram of the minimum distance between the point to be measured and the inner and outer contours of the gun position intelligent detection system based on image-based temperature measurement sampling in an embodiment of the present invention.
  • Fig. 7 is a schematic diagram of the gun position calculated according to the closest point strategy of the image-based temperature measurement sampling lower gun position intelligent detection system in the embodiment of the present invention.
  • Fig. 8 is a schematic diagram of an image-based temperature measurement sampling lower gun position intelligent detection system in an embodiment of the present invention, which is calculated according to the most suitable point strategy.
  • the image-based method for intelligent detection of gun position under temperature measurement and sampling in this embodiment includes:
  • the target image is the image of the molten steel surface
  • the contour information of the steel slag crack in the molten steel surface obtains the contour information of the steel slag crack in the molten steel surface, and the contour information includes the inner and outer contours of the steel slag crack;
  • the preset temperature measurement sampling gun position, the area of interest around the gun position, and the smallest area of the gun are adapted to the obtained steel slag gap width, and the gun position that meets the conditions of the gun is selected from the molten steel level Take temperature sampling.
  • the temperature measurement sampling gun position, the region of interest around the gun position, and the minimum area of the gun are preset in the image, and the minimum area of the gun is used as the gun down condition.
  • the condition is that when the gun head is inserted into the molten steel surface, there is no steel slag in the neighborhood around the molten steel surface that the gun head contacts.
  • the minimum area size in this embodiment is 10 ⁇ 10 cm by default.
  • the manipulator is a six-axis manipulator.
  • the image acquisition system can be installed on the flange at the end of the six-axis manipulator, and the manipulator can be taught to the ladle detection position, and the original path of the action is recorded.
  • the manipulator moves to the ladle detection position and collects a new ladle steel liquid level image as the detection target image.
  • the dynamic threshold binarization operation is performed on the region of interest of the molten steel surface image, and then the connected domain detection is performed on the binarized image, and the detection result is subjected to region-based morphological processing to obtain the steel slag crack
  • the inner and outer contours; the distance between the inner and outer contours of the steel slag gap is the width of the steel slag gap.
  • Adopting the nearest point strategy refers to calculating the minimum area of a certain point in the neighborhood around the position of the gun that satisfies the gun.
  • this strategy is to meet the conditions of the lower gun; the strategy of adopting the most suitable point means to calculate the area with the largest width of the steel slag gap in the area of interest around the lower gun position as the lower gun Position, even when the steel liquid level is dynamically changing, the width of the steel slag gap in this area still meets the lower gun conditions.
  • the gray-scale histogram analysis is performed on the region of interest of the steel liquid level image, and the low-tail and high-tail limits of the gray-scale histogram are counted.
  • the area is considered to be steel slag, and the area is determined to be a steel slag area.
  • the gray value in the area is higher than the high tail value, the area is considered to be molten steel, and the area is determined to be a molten steel area.
  • the molten steel level contains both slag and molten steel, and the area is determined to be a mixed zone of both molten steel and molten steel on the molten steel surface.
  • the degree histogram performs a dynamic threshold binarization operation on the region to obtain a binarized image. Then analyze the connected domain of the binarized image, and the black area obtained is the steel slag area, and the white area obtained is the steel slag crack area. By searching the edges of the steel slag gap area, the inner and outer contours of the area are obtained, and the distance between the inner and outer contours is defined as the width of the steel slag gap.
  • the shooting strategy can be formulated in advance, and one of the closest point strategy or the most suitable point strategy can be selected for temperature measurement sampling. It can also include adopting the most suitable point when the closest point strategy is used to find a location that meets the conditions and fails.
  • the strategy determines the position of the gun and conducts temperature measurement and sampling. For example, the nearest point strategy is used to find all the steel slag gap widths in the neighborhood around the lower gun position, and the position that meets the minimum area condition of the lower gun is calculated. When the nearest point strategy is used to find the location that meets the conditions, the most suitable point strategy is used to find the sense. The positions of all the steel slag gaps that meet the conditions in the area of interest and the width of the steel slag gaps corresponding to this position. The position with the largest width value is the most suitable point for gun down.
  • this embodiment also provides an image-based system for intelligent detection of gun position under temperature measurement and sampling, including:
  • the image acquisition module is used to acquire the target image, the target image is the image of the molten steel surface;
  • the image processing module is used to obtain the profile information of the steel slag gap in the molten steel surface according to the collected target image.
  • the profile information includes the inner and outer contours of the steel slag gap; according to the profile information of the steel slag gap, the width of the steel slag gap is obtained and the preset Temperature measurement and sampling of the lower gun position, the area of interest around the lower gun position and the smallest area of the lower gun, to adapt to the obtained steel slag gap width;
  • the temperature measurement and sampling module is used to select the lower gun position that meets the lower gun conditions in the steel liquid level for temperature measurement and sampling.
  • the image acquisition module may include image acquisition components such as industrial cameras, industrial lenses, filters, etc.
  • the image acquisition module is fixed to the mobile end of the temperature measurement and sampling manipulator through the mounting bracket to collect target images.
  • the image processing module can use a technical industrial computer to obtain the contour information of the steel slag gap in the molten steel surface according to the collected target image.
  • the contour information includes the inner and outer contours of the steel slag gap; obtain the width of the steel slag gap according to the contour information of the steel slag gap .
  • the temperature measurement and sampling module is used as an actuator to carry out the front-end work of temperature measurement and sampling.
  • the image acquisition protection module includes a cooling unit for reducing the temperature of the working environment of the image acquisition system and protection for heat insulation, radiation, and dust.
  • the cooling unit can use air cooling to reduce the working environment temperature of the image acquisition module.
  • the installation module includes a mounting bracket and a temperature measurement and sampling manipulator.
  • the alarm module can use sound and light alarm lights and other warning devices. When the gun is detected An alarm is issued when the position fails.
  • the communication module also includes a communication module for data transmission between the image acquisition module and the image processing module.
  • the communication module may adopt a wired communication module, such as communication through the TCP/IP network protocol.
  • the communication module It can be installed at the end of the temperature measurement and sampling robot. After the robot moves to the detection position, the image collected by the image acquisition module is transmitted to the image processing module for detection, and the detection result is sent to the robot.
  • system control process includes:
  • the signal is fed back to the image processing module, and the image processing module triggers the image acquisition module to collect the ladle liquid level image and use it as the target image;
  • the region of interest around the position of the gun, the position of the gun for temperature measurement in the image, and the minimum area of the gun are preset.
  • the minimum area of the gun is used as the gun down condition. The condition is that when the gun head is inserted into the molten steel surface, there is no steel slag in the neighborhood around the molten steel surface that the gun head touches.
  • the gray-scale histogram analysis is performed on the region of interest of the steel liquid level image, and the low-tail and high-tail limits of the gray-scale histogram statistics are limited, and the gray values in the region are all lower than the low-tail value 1
  • the gray value in the region is higher than the high tail value of 3
  • the molten steel surface contains both slag and molten steel.
  • the dynamic threshold binarization operation is performed on the region through the gray histogram, and the dynamic threshold value 2 is used to obtain Binarize the image.
  • the connected domain analysis is performed on the binarized image, and the obtained black area is the steel slag area, and the obtained white area is the steel slag crack area, as shown in FIG. 4.
  • the edge search is performed on the steel slag gap area to obtain the inner and outer contours of the area.
  • the black and white colors of the picture and can be distinguished by different colors in practical applications, for example,
  • the red outline is the outer outline of the steel slag seam
  • the blue outline is the inner outline of the steel slag seam
  • the distance between the inner and outer outline is defined as the width of the steel slag seam.
  • point A is the point to be measured
  • 61 is the region of interest
  • 62 is the outer contour
  • 63 and 64 are the inner contours.
  • Select the point to be measured in the region of interest around the lower gun position and determine the relationship between the point to be measured and the outer contour (62) and the inner contour (63 and 64) to determine whether the point A to be measured is in the steel slag crack.
  • the judgment rule is: the point A to be measured is both inside the outer contour (62) and outside the inner contour (63 and 64), then the point A to be measured is the point inside the steel slag gap.
  • the nearest point strategy is used to find all the slag gap widths in the neighborhood around the lower gun position, and the position that meets the minimum area condition of the lower gun is calculated.
  • the circle mark is the preset gun lower position.
  • the size of the neighborhood around the gun position can be preset. Calculate the minimum distance between each point in the steel slag gap and the inner and outer contours in the neighborhood around the gun position, compare the distance of each point, and find the maximum value of all distances, as shown by the cross mark, when the maximum value is greater than When the gun is in the smallest area, it means that there are gun dropping points that meet the closest point strategy in the surrounding neighborhood of the preset gun dropping area.
  • the most suitable strategy is used to find all the widths of steel slag gaps in the region of interest in the lower gun position, and the position that satisfies the minimum area condition of the lower gun is calculated.
  • the method is the same as the closest point strategy. The only difference is that the closest point strategy only The points in the neighborhood around the lower gun position are traversed as the points to be measured, and the most suitable point strategy uses all the points in the region of interest of the lower gun position as the points to be measured.
  • the calculation results are shown in Figure 8.
  • the cross mark in the figure is the position of the lower gun point calculated by the most suitable point strategy.
  • the position of the gun point is calculated and the coordinates are sent to the manipulator. If the calculation fails, an alarm will be issued.
  • the manipulator moves to the new gun point to perform temperature measurement and sampling operations.

Landscapes

  • Engineering & Computer Science (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Radiation Pyrometers (AREA)

Abstract

一种基于图像的测温取样下枪位置智能检测方法及系统，方法包括：采集目标图像，目标图像为钢液面图像；根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，轮廓信息包括钢渣缝的内轮廓（63、64）和外轮廓（62）；根据钢渣缝轮廓信息，获取钢渣缝宽度；将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，在钢液面中选取满足下枪条件的下枪位置进行测温取样；通过图像检测钢包的钢液面上钢渣的分布情况，计算出所预设的下枪区域内所有的钢渣缝宽度，并通过预定的策略选取满足下枪条件的下枪位置，进行测温取样，从而提高了机械手测温取样作业的智能化水平。

Description

一种基于图像的测温取样下枪位置智能检测方法及系统 技术领域
本发明涉及冶金领域，尤其涉及一种基于图像的测温取样下枪位置智能检测方法及系统。
背景技术
在冶金领域中，提高炼钢的自动化水平，可以增强整个设备操作的安全性和可靠性，对保障炼钢钢水质量，提高炼钢的劳动生产率起着至关重要的作用。作为炼钢过程工艺操作的依据，需要在炼钢出钢后对钢水进行测温及取样，现有技术中通常采用通过机械手来进行测温取样。
但是，机械手在钢包精炼炉测温取样的实际应用中，通常使用固定路径来完成作业，然而对于炼钢作业工艺流程中，由于钢液面上通常有钢渣，机械手使用固定路径测温取样时，枪头可能会因碰撞到钢渣而损坏枪头，导致无法满足测温取样的要求，现有的技术中，没有针对钢渣缝宽度来调整下枪位置的检测方法，综上所述，需要一种新的检测方法检测下枪点区域内钢渣的分布情况，并精确定位适合下枪的钢渣缝区域。
发明内容
鉴于以上所述现有技术的缺点，本发明提供一种基于图像的测温取样下枪位置智能检测方法及系统，以解决上述技术问题。
本发明提供的基于图像的测温取样下枪位置智能检测方法，包括：
采集目标图像，所述目标图像为钢液面图像；
根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，所述轮廓信息包括钢渣缝的内轮廓和外轮廓；
根据所述钢渣缝轮廓信息，获取钢渣缝宽度；
将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，在钢液面中选取满足下枪条件的下枪位置进行测温取样。
可选的，对获取所述目标图像的感兴趣区域，进而获取所述目标图像的感兴趣区域的灰度直方图，根据预先设置的灰度阈值，获取所述钢渣缝轮廓信息。
可选的，所述灰度阈值包括低尾部值和高尾部值，
当区域内的灰度值低于所述低尾部值时，则判定该区域内为钢渣区；
当区域内的灰度值高于所述高尾部值时，则判定该区域内为钢液区；
当区域内的灰度值处于低尾部值与高尾部值之间时，则判定该区域内为钢液面既有钢渣也有钢液的混合区。
可选的，对所述混合区动态阈值二值化处理，获取二值化图像，根据所述二值化图像获取混合区内的钢渣区域和钢渣缝区域。
可选的，对所述钢渣缝区域进行边缘查找，获取所述钢渣缝区域的内轮廓和外轮廓，并将内轮廓和外轮廓之间的距离作为钢渣缝宽度。
可选的，将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，根据预设的下枪策略进行测温取样，所述预设的下枪策略包括最近点策略和/或最适合点策略，其中，
最近点策略包括选取预设的测温取样下枪位置周围邻域内中满足下枪的最小区域作为下枪位置；
最适合点策略包括选取预设的测温取样下枪位置感兴趣区域内钢渣缝宽度最大的区域作为下枪位置。
可选的，所述预设的下枪策略还包括当采用最近点策略查找满足条件的位置失败时，采用最适合点的策略确定下枪位置，进行测温取样。
本发明还提供一种基于图像的测温取样下枪位置智能检测系统，包括
图像采集模块，用于采集目标图像，所述目标图像为钢液面图像；
图像处理模块，用于根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，所述轮廓信息包括钢渣缝的内轮廓和外轮廓；根据所述钢渣缝轮廓信息，获取钢渣缝宽度；
测温取样模块，用于将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，在钢液面中选取满足下枪条件的下枪位置进行测温取样。
可选的，还包括图像采集防护模块，所述图像采集防护模块包括用于降低图像采集系统工作环境温度的冷却单元和用于隔热防辐射及防尘的防护罩。
可选的，还包括安装模块和用于下枪位置检测失败时进行报警的报警模块；所述安装模块包括安装支架和测温取样机械手，所述图像采集模块通过所述安装支架固定于测温取样机械手的移动端。
本发明的有益效果：本发明中的基于图像的测温取样下枪位置智能检测方法及系统，通过图像检测钢包的钢液面上钢渣的分布情况，计算出所预设的下枪区域内所有的钢渣缝宽度，并通过预定的策略选取满足下枪条件的下枪位置，进行测温取样，从而提高了机械手测温取样作业的智能化水平。
附图说明
图1是本发明实施例中基于图像的测温取样下枪位置智能检测方法的流程示意图。
图2是本发明实施例中基于图像的测温取样下枪位置智能检测系统的结构示意图。
图3是本发明实施例中基于图像的测温取样下枪位置智能检测系统的目标检测图像及感兴趣区域示意图。
图4是本发明实施例中基于图像的测温取样下枪位置智能检测系统的感兴趣区域的灰度直方图。
图5是本发明实施例中基于图像的测温取样下枪位置智能检测系统的感兴趣区域二值图像内外轮廓图。
图6是本发明实施例中基于图像的测温取样下枪位置智能检测系统的待测点与内外轮廓的距离最小值示意图。
图7是本发明实施例中基于图像的测温取样下枪位置智能检测系统的根据最近点策略计算得到的下枪点示意图。
图8是本发明实施例中基于图像的测温取样下枪位置智能检测系统的根据最适合点策略计算得到的下枪点示意图。
具体实施方式
以下通过特定的具体实例说明本发明的实施方式，本领域技术人员可由本说明书所揭露的内容轻易地了解本发明的其他优点与功效。本发明还可以通过另外不同的具体实施方式加以实施或应用，本说明书中的各项细节也可以基于不同观点与应用，在没有背离本发明的精神下进行各种修饰或改变。需说明的是，在不冲突的情况下，以下实施例及实施例中的特征可以相互组合。
需要说明的是，以下实施例中所提供的图示仅以示意方式说明本发明的基本构想，遂图式中仅显示与本发明中有关的组件而非按照实际实施时的组件数目、形状及尺寸绘制，其实际实施时各组件的型态、数量及比例可为一种随意的改变，且其组件布局型态也可能更为复杂。
在下文描述中，探讨了大量细节，以提供对本发明实施例的更透彻的解释，然而，对本领域技术人员来说，可以在没有这些具体细节的情况下实施本发明的实施例是显而易见的，在其他实施例中，以方框图的形式而不是以细节的形式来示出公知的结构和设备，以避免使本发明的实施例难以理解。
如图1所示，本实施例中的基于图像的测温取样下枪位置智能检测方法，包括：
采集目标图像，目标图像为钢液面图像；
根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，轮廓信息包括钢渣缝的内轮廓和外轮廓；
根据所述钢渣缝轮廓信息，获取钢渣缝宽度；
将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，在钢液面中选取满足下枪条件的下枪位置进行测温取样。
在本实施例中，预先设置图像中测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，将下枪的最小区域作为下枪条件，本实施例中的下枪条件指枪头插入钢液面时，枪头所接触的钢液面周围邻域内无钢渣，可选的，本实施例中的最小区域大小缺省值为10X10cm。
可选的，在本实施例中，机械手为六轴机械手，可以将图像采集系统安装于六轴机械手末端的法兰盘上，通过示教机械手到钢包检测位，并记录动作的原始路径，当机械手运动到钢包检测位，采集新的钢包钢液面图像作为检测的目标图像。
在本实施例中，通过对钢液面图像的感兴趣区域进行动态阈值二值化运算，再对二值化图像进行连通域检测，并将检测结果进行基于区域的形态学处理，得到钢渣缝内轮廓和外轮廓；钢渣缝内轮廓和外轮廓之间的距离即为钢渣缝宽度。采用最近点策略是指计算下枪位置周围邻域内某点满足下枪的最小区域。对于处于微小变化状态的钢液面，通过这种策略是满足下枪条件的；采用最适合点的策略是指，计算出下枪位置周围的感兴趣区域内钢渣缝宽度最大的区域作为下枪位置，即使钢液面处于动态变化时，该区域的钢渣缝 宽度仍然满足下枪条件。
在本实施例中，对钢液面图像的感兴趣区域进行灰度直方图分析，并对灰度直方图统计的低尾部和高尾部限制，区域内的灰度值都低于低尾部值时，则认为区域内都为钢渣，判定该区域内为钢渣区，区域内的灰度值都高于高尾部值时，则认为区域内都为钢液，判定该区域内为钢液区。区域内的灰度值处于低尾部值与高尾部值之间时，则钢液面既有钢渣也有钢液，则判定该区域内为钢液面既有钢渣也有钢液的混合区，通过灰度直方图对该区域进行动态阈值二值化运算，得到二值化图像。再对二值化图像进行连通域分析，得到的黑色区域为钢渣区域，得到的白色区域为钢渣缝区域。通过对钢渣缝区域进行边缘查找，得到该区域的内轮廓和外轮廓，并定义内轮廓和外轮廓之间的距离即为钢渣缝宽度。
在本实施例中，下枪策略可以预先制定，选择最近点策略或者最合适点策略之一进行测温取样，也可以包括当采用最近点策略查找满足条件的位置失败时，采用最适合点的策略确定下枪位置，进行测温取样。如采用最近点策略查找下枪位置周围邻域内所有的钢渣缝宽度，计算满足下枪最小区域条件的位置，当采用最近点策略查找满足条件的位置失败时，再采用最适合点的策略查找感兴趣区域内所有满足条件的的钢渣缝区域的位置和该位置对应的钢渣缝宽度，最大宽度值的位置即为最适合的下枪点。
相应地，本实施例还提供一种基于图像的测温取样下枪位置智能检测系统，包括：
图像采集模块，用于采集目标图像，目标图像为钢液面图像；
图像处理模块，用于根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，轮廓信息包括钢渣缝的内轮廓和外轮廓；根据钢渣缝轮廓信息，获取钢渣缝宽度，将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配；
测温取样模块，用于在钢液面中选取满足下枪条件的下枪位置进行测温取样。
在本实施例中，图像采集模块可以包括工业相机、工业镜头、滤光片等图像采集部件，图像采集模块通过所述安装支架固定于测温取样机械手的移动端，对目标图像进行采集。图像处理模块可以采用技术中的工控机，根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，轮廓信息包括钢渣缝的内轮廓和外轮廓；根据钢渣缝轮廓信息，获取钢渣缝宽度。测温取样模块作为执行机构，进行测温取样的前端工作。
在本实施例中，还包括图像采集防护模块、安装模块和报警模块，其中，图像采集防护模块包括用于降低图像采集系统工作环境温度的冷却单元和用于隔热防辐射及防尘的防护罩，冷却单元可以采用气冷的方式来降低图像采集模块的工作环境温度，安装模块包括安装支架和测温取样机械手，报警模块可以采用声光报警灯等具有警示作用的器件，当检测下枪位置失败时进行报警。
在本实施例中，还包括通信模块，用于图像采集模块和图像处理模块之间的数据传输，通信模块可以采用有线通信模块，例如通过TCP/IP网络协议进行通信，可选的，通信模块可以安装于测温取样机器人末端，机器人移动到检测位后，将图像采集模块采集的图像传输给图像处理模块进行检测，将检测结果发送给机器人。
在本实施例中，系统控制流程包括：
S11、系统收到测温取样命令后，控制机械手运动到测温取样检测位；
S12、机械手运动到检测位后，反馈信号给图像处理模块，图像处理模块触发图像 采集模块采集钢包液面图像，并作为目标图像；
S13、在图像处理模块中预先设置最近点和/或最适合点的策略，并分别设置两种策略不同的优先级，按照优先级由高到低地通过测温取样模块执行。
在本实施例中，预先设置下枪位置周围的感兴趣区域、图像中测温取样下枪位置和下枪的最小区域，如图3所示，下枪的最小区域作为下枪条件，下枪条件指枪头插入钢液面时，枪头所接触的钢液面周围邻域内无钢渣。
在本实施例中，对钢液面图像的感兴趣区域进行灰度直方图分析，并对灰度直方图统计的低尾部和高尾部限制，区域内的灰度值都低于低尾部值1时，则认为区域内都为钢渣，区域内的灰度值都高于高尾部值3时，则认为区域内都为钢液。区域内的灰度值处于低尾部值与高尾部值之间时，则钢液面既有钢渣也有钢液，通过灰度直方图对该区域进行动态阈值二值化运算，利用动态阈值2得到二值化图像。在本实施例中，对二值化图像进行连通域分析，得到的黑色区域为钢渣区域，得到的白色区域为钢渣缝区域，如图4所示。
在本实施例中，对钢渣缝区域进行边缘查找，得到该区域的内轮廓和外轮廓，如图5所示，受限于图片黑白色彩，在实际应用中可以通过不同颜色来进行区分，例如采用红色的轮廓为钢渣缝的外轮廓，蓝色的轮廓为钢渣缝的内轮廓，并定义内轮廓和外轮廓之间的距离即为钢渣缝宽度。
具体地，首先需要查找钢渣缝的点，查找方法如下：
如图6所示，A点为待测点，61为感兴趣区域，62为外轮廓，63和64为内轮廓。在下枪位置周围的感兴趣区域内选择待测点A，判断待测点A与外轮廓(62)与内轮廓(63和64)的关系从而确定待测点A是否在钢渣缝内。判断规则为：待测点A既在外轮廓(62)内部，又在内轮廓(63和64)外部，则待测点A为钢渣缝内的点。计算钢渣缝内的待测点A距离外轮廓(62)和内轮廓(63和64)的最小距离，定义为待测点的钢渣缝宽度。
在本实施例中，如图7所示，采用最近点策略查找下枪位置周围邻域内所有的钢渣缝宽度，计算满足下枪最小区域条件的位置，圈形标记为预先设置的下枪位置，下枪位置周围邻域大小可预先设置。计算下枪位置周围邻域内每个属于钢渣缝内的点与内外轮廓的最小距离，将每个点的距离进行比较，求出所有距离的最大值，如十字标记所示，当该最大值大于最小的下枪区域时，则表示预先设置的下枪区域的周围邻域内有满足最近点策略的下枪点。
在本实施例中，采用最适合策略查找下枪位置感兴趣区域内所有的钢渣缝宽度，计算满足下枪最小区域条件的位置，方法同最近点策略，唯一的区别在于，最近点策略中只遍历下枪位置周围邻域的点作为待测点，而最适合点策略使用下枪位置的感兴趣区域内所有的点作为待测点，计算结果如图8所示。图中十字标记为最适合点策略计算得到的下枪点位置。
根据两种策略智能检测后，计算出下枪点位置，并将坐标发送给机械手，若计算失败则报警。
然后，机械手运动到新的下枪点，进行测温取样作业。
上述实施例仅例示性说明本发明的原理及其功效，而非用于限制本发明。任何熟悉此技术的人士皆可在不违背本发明的精神及范畴下，对上述实施例进行修饰或改变。因此，举凡所属技术领域中具有通常知识者在未脱离本发明所揭示的精神与技术思想下所完 成的一切等效修饰或改变，仍应由本发明的权利要求所涵盖。

Claims (10)

  1. 一种基于图像的测温取样下枪位置智能检测方法，其特征在于，包括：
    采集目标图像，所述目标图像为钢液面图像；
    根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，所述轮廓信息包括钢渣缝的内轮廓和外轮廓；
    根据所述钢渣缝轮廓信息，获取钢渣缝宽度；
    将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，在钢液面中选取满足下枪条件的下枪位置进行测温取样。
  2. 根据权利要求1所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，对获取所述目标图像的感兴趣区域，进而获取所述目标图像的感兴趣区域的灰度直方图，根据预先设置的灰度阈值，获取所述钢渣缝轮廓信息。
  3. 根据权利要求2所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，所述灰度阈值包括低尾部值和高尾部值，
    当区域内的灰度值低于所述低尾部值时，则判定该区域内为钢渣区；
    当区域内的灰度值高于所述高尾部值时，则判定该区域内为钢液区；
    当区域内的灰度值处于低尾部值与高尾部值之间时，则判定该区域内为钢液面既有钢渣也有钢液的混合区。
  4. 根据权利要求3所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，对所述混合区动态阈值二值化处理，获取二值化图像，根据所述二值化图像获取混合区内的钢渣区域和钢渣缝区域。
  5. 根据权利要求4所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，对所述钢渣缝区域进行边缘查找，获取所述钢渣缝区域的内轮廓和外轮廓，并将内轮廓和外轮廓之间的距离作为钢渣缝宽度。
  6. 根据权利要求1所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配，根据预设的下枪策略进行测温取样，所述预设的下枪策略包括最近点策略和/或最适合点策略，其中，
    最近点策略包括选取预设的测温取样下枪位置周围邻域内中满足下枪的最小区域作为下枪位置；
    最适合点策略包括选取预设的测温取样下枪位置感兴趣区域内钢渣缝宽度最大的区域作为下枪位置。
  7. 根据权利要求6所述的基于图像的测温取样下枪位置智能检测方法，其特征在于，所述预设的下枪策略还包括当采用最近点策略查找满足条件的位置失败时，采用最适合点的策略确定下枪位置，进行测温取样。
  8. 一种基于图像的测温取样下枪位置智能检测系统，其特征在于，包括
    图像采集模块，用于采集目标图像，所述目标图像为钢液面图像；
    图像处理模块，用于根据采集的目标图像，获取钢液面中的钢渣缝轮廓信息，所述轮廓信息包括钢渣缝的内轮廓和外轮廓；根据所述钢渣缝轮廓信息，获取钢渣缝宽度，将预设的测温取样下枪位置、下枪位置周围的感兴趣区域和下枪的最小区域，与获取的钢渣缝宽度进行适配；
    测温取样模块，用于在钢液面中选取满足下枪条件的下枪位置进行测温取样。
  9. 根据权利要求8所述的基于图像的测温取样下枪位置智能检测系统，其特征在于，还包括图像采集防护模块，所述图像采集防护模块包括用于降低图像采集系统工作环境温度的冷却单元和用于隔热防辐射及防尘的防护罩。
  10. 根据权利要求9所述的基于图像的测温取样下枪位置智能检测系统，其特征在于，还包括安装模块和用于下枪位置检测失败时进行报警的报警模块；所述安装模块包括安装支架和测温取样机械手，所述图像采集模块通过所述安装支架固定于测温取样机械手。
PCT/CN2020/112190 2020-04-16 2020-08-28 一种基于图像的测温取样下枪位置智能检测方法及系统 WO2021208324A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
DE112020000297.7T DE112020000297T5 (de) 2020-04-16 2020-08-28 Bildbasiertes Verfahren und System zur intelligenten Erkennung der Pistolenauslöseposition zur Temperaturmessung und Probenahme

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010297507.3A CN111397766A (zh) 2020-04-16 2020-04-16 一种基于图像的测温取样下枪位置智能检测方法及系统
CN202010297507.3 2020-04-16

Publications (1)

Publication Number Publication Date
WO2021208324A1 true WO2021208324A1 (zh) 2021-10-21

Family

ID=71429556

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/112190 WO2021208324A1 (zh) 2020-04-16 2020-08-28 一种基于图像的测温取样下枪位置智能检测方法及系统

Country Status (3)

Country Link
CN (1) CN111397766A (zh)
DE (1) DE112020000297T5 (zh)
WO (1) WO2021208324A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115307783A (zh) * 2022-07-29 2022-11-08 北京瓦特曼智能科技有限公司 用于钢水测温取样系统的方法和控制器
CN116337861A (zh) * 2023-03-27 2023-06-27 广州市倍乐食品有限公司 卡达酱加热控制系统及方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111397766A (zh) * 2020-04-16 2020-07-10 中冶赛迪工程技术股份有限公司 一种基于图像的测温取样下枪位置智能检测方法及系统
CN114088157B (zh) * 2021-11-18 2024-05-07 中冶赛迪技术研究中心有限公司 钢水液面检测方法、设备及介质
CN115537493B (zh) * 2022-09-23 2024-03-29 山东钢铁股份有限公司 一种铁水测温取样机器人及其操控方法

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010121937A (ja) * 2008-11-17 2010-06-03 Jfe Steel Corp 溶融物のサンプリング方法および装置
CN102392098A (zh) * 2011-11-04 2012-03-28 衡阳镭目科技有限责任公司 一种炼钢炉下渣检测系统
CN204128857U (zh) * 2014-09-19 2015-01-28 中国重型机械研究院股份公司 一种全自动炉外测温取样装置
CN109029737A (zh) * 2018-09-05 2018-12-18 中冶赛迪技术研究中心有限公司 一种测温取样的智能视觉定位方法
CN109146874A (zh) * 2018-09-05 2019-01-04 中冶赛迪技术研究中心有限公司 一种基于机器视觉的溢渣检测方法
CN110728712A (zh) * 2019-09-26 2020-01-24 中冶赛迪工程技术股份有限公司 基于机器视觉的炉口定位方法、装置、存储介质及电子终端
CN110751669A (zh) * 2019-09-24 2020-02-04 武汉科技大学 一种新型cbocp在线红外转炉出钢钢流自动检测与跟踪方法及系统
CN111397766A (zh) * 2020-04-16 2020-07-10 中冶赛迪工程技术股份有限公司 一种基于图像的测温取样下枪位置智能检测方法及系统

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010121937A (ja) * 2008-11-17 2010-06-03 Jfe Steel Corp 溶融物のサンプリング方法および装置
CN102392098A (zh) * 2011-11-04 2012-03-28 衡阳镭目科技有限责任公司 一种炼钢炉下渣检测系统
CN204128857U (zh) * 2014-09-19 2015-01-28 中国重型机械研究院股份公司 一种全自动炉外测温取样装置
CN109029737A (zh) * 2018-09-05 2018-12-18 中冶赛迪技术研究中心有限公司 一种测温取样的智能视觉定位方法
CN109146874A (zh) * 2018-09-05 2019-01-04 中冶赛迪技术研究中心有限公司 一种基于机器视觉的溢渣检测方法
CN110751669A (zh) * 2019-09-24 2020-02-04 武汉科技大学 一种新型cbocp在线红外转炉出钢钢流自动检测与跟踪方法及系统
CN110728712A (zh) * 2019-09-26 2020-01-24 中冶赛迪工程技术股份有限公司 基于机器视觉的炉口定位方法、装置、存储介质及电子终端
CN111397766A (zh) * 2020-04-16 2020-07-10 中冶赛迪工程技术股份有限公司 一种基于图像的测温取样下枪位置智能检测方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115307783A (zh) * 2022-07-29 2022-11-08 北京瓦特曼智能科技有限公司 用于钢水测温取样系统的方法和控制器
CN115307783B (zh) * 2022-07-29 2024-02-06 北京瓦特曼智能科技有限公司 用于钢水测温取样系统的方法和控制器
CN116337861A (zh) * 2023-03-27 2023-06-27 广州市倍乐食品有限公司 卡达酱加热控制系统及方法
CN116337861B (zh) * 2023-03-27 2023-10-13 广州市倍乐食品有限公司 卡达酱加热控制系统及方法

Also Published As

Publication number Publication date
CN111397766A (zh) 2020-07-10
DE112020000297T5 (de) 2022-03-17

Similar Documents

Publication Publication Date Title
WO2021208324A1 (zh) 一种基于图像的测温取样下枪位置智能检测方法及系统
CN110480127B (zh) 一种基于结构光视觉传感的焊缝跟踪系统及方法
CN109447011B (zh) 红外对蒸汽管道泄露的实时监控方法
CN105353654B (zh) 一种基于图像处理的铁水扒渣检测与控制系统及其方法
CN105160683B (zh) 一种基于人工干预的铁水扒渣检测与控制系统及其方法
CN106441584A (zh) 基于红外测温的转炉下渣检测方法
CN107610085A (zh) 一种基于计算机视觉的焊点缺陷检测系统
CN111445444B (zh) 一种基于极化特征的铁水流速检测方法
WO2021088247A1 (zh) 合金分析视觉定位方法、装置及合金分析系统
CN109117764A (zh) 在电力监控中采用颜色阈值法识别目标物体区域电气符号的方法
CN105562643A (zh) 一种结晶器漏钢处理方法
CN109409395A (zh) 在电力监控中采用模板匹配法识别目标物体区域电气符号的方法
CN102455171A (zh) 一种激光拼焊焊缝背面几何形貌检测方法及其实现装置
CN110751669A (zh) 一种新型cbocp在线红外转炉出钢钢流自动检测与跟踪方法及系统
KR20140056790A (ko) 영상 인식 장치 및 그 방법
CN111639568A (zh) 一种用于变电站智能机器人的导航巡视方法
CN110378937B (zh) 基于Kinect摄像头的工业机械臂人机安全距离检测方法
CN111560489A (zh) 一种转炉出钢中的台车自动走行方法和系统
CN205193472U (zh) 一种基于图像处理的铁水扒渣检测与控制系统
JP7245877B2 (ja) 画像に基づく測温サンプリングのためのマニピュレータ入れ位置のインテリジェント検出方法及びシステム
CN112489125A (zh) 一种堆场行人自动检测方法及装置
CN109976351B (zh) 一种电力巡检机器人定点定位系统
JPH11345392A (ja) 障害物検出装置及び障害物検出方法
Zhang et al. A study on seam tracking in robotic GMA welding process
CN111203655A (zh) 一种激光焊接过程三维空间飞溅识别装置

Legal Events

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

Ref document number: 20931173

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 20931173

Country of ref document: EP

Kind code of ref document: A1