WO2012171627A1

WO2012171627A1 - Method and device of determining a tracking characteristic and/or strip width of a moving strip

Info

Publication number: WO2012171627A1
Application number: PCT/EP2012/002456
Authority: WO
Inventors: Erwin Johannes Geradus SPELBOS
Original assignee: Tata Steel Nederland Technology B.V.
Priority date: 2011-06-14
Filing date: 2012-06-11
Publication date: 2012-12-20

Abstract

The present invention relates among others to a method of determining a tracking characteristic of a moving strip, e.g. a metal strip in a mill facility, comprising an edge determination step using vision means, wherein repeatedly images of the moving strip are taken and the images are evaluated using image processing thereby determining the longitudinal edge positions of the strip, from which edge positions the tracking characteristic is determined. According to the invention, the vision means comprises an area scan camera for scanning a predetermined area covering at least the width of the strip and regions adjacent each side edge over a certain length of the moving strip. The invention also relates to a method of steering a moving strip, a device for determining a tracking characteristic of a moving strip, and a mill facility for rolling a metal strip comprising such a device.

Description

METHOD AND DEVICE OF DETERMINING A TRACKING

CHARACTERISTIC AND/OR STRIP WIDTH OF A MOVING STRIP

The present invention relates to a method of determining a tracking characteristic selected from the group comprising strip centre line position and orientation, camber, and optionally strip width, of a moving strip, in particular a moving metal strip in a mill facility, as well as a method of steering a moving strip, a device for determining strip centre line position and orientation and camber of a moving strip and a mill facility provided with such a device.

A metal strip such as steel strip is usually manufactured in a steel rolling mill from a cast slab. In the mill the cast slab is forced through pairs of rollers with increasingly smaller gaps. As the steel strip passes through the subsequently arranged pairs of rollers, the steel strip may have a tendency for undesired lateral movement involving the risk of collision with the mill and commonly employed side-guides. Collision would usually result in undesirable and expensive maintenance and downtime, reduction of strip quality and mill productivity. Usually the strip is steered by the operators who can adjust the tilts of the pairs of rollers. Alternatively, the steering process has been automated using differential force measurements, e.g. using load cells, in the roll stands. The relationship between strip position and differential force is highly non-linear and the complexities of this kind of system can be overcome by using a direct measurement, such as a vision system.

In a known system put in practice three line scan cameras are arranged at positions spaced apart along the moving strip. From the image lines obtained the longitudinal edge positions can be derived, which positions can be used for calculating other characteristic values of the moving metal strip. These characteristic values can be used in a steering process. A disadvantage of this known method is that it is very complicated to correlate a given data set for a certain strip position at one point in time to a further data set thereof at another subsequent point in time. Another disadvantage is that the images taken may be seriously distorted or even be unusable due to the presence of steam, water and dust. Then reliable data cannot be obtained.

"The development of a vision-based system for strip position measurement during hot strip rolling" by Ben Carruthers-Watt, Cardiff University, describes a camera-based measurement system sensitive to the visible light spectrum and image analysis software, suitable for overcoming the challenges of operating in the noisy and dirty environment of steelworks, in particular a vision system with control on the first two stands of the finishing train in a hot mill. The cameras are arranged at a position perpendicular to the moving strip. The images taken are fed to a computing system that performs an image analysis using Bezier curves in order to determine the position of the side edges of the strip and then calculates the location of the strip centreline as close to the rolls as possible. The thus obtained data can be used to control an automated steering unit for generating a lateral movement of the strip, if needed.

A disadvantage of this known system is that it is still hard to detect the edges in the images in a reliable manner because it is based on brightness differences in the visual spectrum. Furthermore in this solution, the obtained images cover a fixed and relatively small area of the strip.

"A machine vision measurement of slab camber in hot strip rolling" by R.J. Montague et all., Journal of Materials Processing Technology 168 (2005), 172-180 also discloses a camera based system for measuring slab camber. A digital camera equipped with a CCD array is mounted on a tripod and positioned on a walkway overlooking the transfer tables at the exit of the first roughing stand. Images of the running slab taken in the visible light spectrum, are subjected to image processing at a later time by firstly performing a projective transformation, a pre-processing step in order to remove the effects of perspective using joints between the transfer table edge guides as control points. The initial image segmentation is carried out using greyscale tresholding techniques. Additional image processing steps are required to deal with sub-optimal conditions, in particular when sunlight spots appear in an image. Drawbacks of this known system are again the relatively poor detection of the slab edges due to the relatively small differences in the visual spectrum between slab and background. Furthermore, pre-processing is performed on the whole raw image, which requires a high computing power and is time-consuming. In fact, it coulc be shown that processing time is too long for allowing adequate steering of the slab based on the obtained results.

An object of the present invention is to provide a method and device for determining the side edges of a moving strip, in particular a moving metal strip, especially a hot steel strip, in a hot mill facility, with high accuracy in order to derive at a characteristic value of the strip such as strip centre line position, strip straightness, camber and strip width. Yet another object of the present invention is to provide a robust method and device for determining the side edges of such a moving strip, which is suitable for use under severe operating conditions, such as those prevailing in a hot mill facility.

Still another object of the present invention is to provide such a real-time method and device resulting in accurate data for steering a moving strip.

According to a first aspect of the invention, one or more of the objects are achieved by a method of determining a tracking characteristic selected from the group comprising strip centre line position and orientation, camber, and optionally strip width, of a moving strip having a temperature of at least 800° C, in particular a moving metal strip in a mill facility, comprising an edge determination step using vision means, wherein repeatedly images of the moving strip are taken and the images are evaluated using image processing thereby determining the side edge positions of the strip, from which edge positions the tracking characteristic and optionally strip width is determined, wherein the vision means comprises a near infrared area scan camera for scanning a predetermined area covering at least the width of the strip and regions adjacent each side edge over a certain length of the moving strip.

In the method according to the invention a near infrared area scan camera, i.e. a camera detecting radiation in the near infrared region (740-1400 micrometer) of the electromagnetic spectrum and producing an image thereof, is used to take images of the moving strip having a temperature of at least 800 °C. Preferably the moving strip is a hot steel strip having a temperature in the range of 810-1200 °C, the strip being treated in a hot mill facility. A camera of this type is less sensitive to light from the surroundings, e.g. sunlight, than a normal camera operating in the visual spectrum only. If the temperature of the moving strip to be imaged is above 800 °C, then the emitted radiation has a significantly higher intensity than at lower temperatures prevailing in the surrounding. Thus the contrast between the hot strip and the regions adjacent each side edge of the strip is relatively large, allowing for an easier detection of the side edges. The camera scans the width of the strip including non-strip regions of the direct neighbourhood at each side edge over a certain strip length. Thus each image or frame comprises a number of image lines offering high resolution and real-time analysis. Images taken by the camera are processed using image analysis in order to arrive at the side edges. From the position of the side edges thus obtained various other values, such as strip centre line position and orientation, camber and strip width can be derived. By viewing an area instead of a single line the accuracy of the edge detection can be improved. If the sampling time of the vision means with respect to the moving strip is such that images subsequently taken in time overlap one another, then if one image fails or is unusable for some reason, the subsequent image taken can be used to deduce the lacking data for the previous image. Moreover the influence of the harsh conditions in a hot roll mill, such as temperature, atmosphere comprising steam, on the accuracy of the method can be effectively reduced. Moreover in a single line camera based system it is difficult to deduct from subsequent images whether any difference is due to a lateral positional change of the strip or to a shape difference of the strip. Using an area scan camera allows to make a distinction between a change of shape and a lateral movement.

In the context of this application strip straightness and camber have almost the same meaning, strip straightness can be determined from a relatively small section of the image taken, while determining camber requires the image taken covers a strip length of about 10 metres.

In a preferred embodiment evaluation of the images comprises determining the strip edges based on IR contrast between the strip and the regions adjacent each side edge. Determination of the strip edges based on contrast measurement can be improved by controlling the lighting conditions, also taking into account the temperature of the moving strip. For example slight overexposure may be beneficial because irregularities of the strip surface will disappear, while the side edges still will show sufficient contrast with the mill environment. Preferably the near infrared area scan camera is a CMOS video camera. Compared to a CCD camera a CMOS camera is less sensitive to affecting neighbouring pixels (bleeding), thereby improving accuracy.

In the method according to the invention preferably the positions of the side edges of the relevant part of the strip are determined from the raw image, prior to making any corrections and/or compensations,. Advantageously the latter operations are only performed with the data (pixels) relating to the side edges. Thereby computing power and time can be saved compared to an image processing method as discloses in Montague et al., where a correction for perspective effects is performed with all image data.

Preferably the images are taken in perspective in order to cover a relatively large area, e.g. from a downwardly directed camera position several metres above the moving strip and tilted in the movement direction of the strip. An image taken in this way covers a length of strip of e.g. a few metres (at least 2 metres) to 15 metres or more. From the image or relevant part thereof it is relatively easy to determine, if any, camber and/or the moving web running at oblique angles. In such a situation, it is preferred that during image processing the distortion of the image due to the angle of view of the area scan camera with respect to the moving strip is compensated. Distortion of the image due to the lens with which the camera is equipped is advantageously also compensated for during image processing. Compensation of these type of distortions, both preferably performed only with the data of the detected side edges and in the order of lens compensation followed by perspective compensation, improves the accuracy of the method according to the invention. Instead of an arrangement of the camera in perspective, the camera might be arranged at a high position with respect to the moving strip in order to cover a large strip area, however at the expense of resolution of the total image and at increased costs for maintenance and repair, even if sufficient head space above the mill stands and the like is available.

As the accuracy of the method according to the invention is important, the distance between the camera and the moving strip is a significant factor. Camera position with respect to the moving strip, looper position (commonly looper angle), and strip thickness (especially at the upstream rollers in a mill facility in view of strip width determination) are examples of additional parameters usually inputted into the image analysis step. The position of the vision means may be monitored, e.g. checked periodically during operation or in between treatments of subsequent strips such as steel slabs in a hot mill, and accordingly controlled. Position determination of the camera can be effected using a marker as a reference. An example of a marker is a physical indication, e.g. a unique stationary feature of the conveyor that moves the strip, that can be retrieved by the camera. However, pollution may cause problems regarding its continuous visibility. Another example is a laser mark. A further possibility is to set the camera between runs to a sensitivity different from the normal operating sensitivity in order to calibrate the camera position, e.g. using different exposure conditions (illumination).

Calibration is commonly employed in order to define a reference or calibration plane, while during actual measurement deviations from the calibration plane are compensated for. Autocalibration or autocorrection e.g. a (semi-)continuous determination of the camera position using the marker and in case of deviations within a certain margin, if necessary, corresponding compensations in the calculations, may be part of the method according to the invention.

Preferably the vision means are protected from the harsh conditions in the mill such as heat and moisture, e.g. by enclosing the camera in a (water-cooled) housing, using a heat shielding element.

In another preferred embodiment the sampling intervals of the vision means with respect to the moving strip are such that images subsequently taken in time are synchronized with strip position or strip speed of subsequently processed strips. This allows for an accurate comparison of the characteristics of these strips and for a better control of the milling process.

A preferred technical application concerns (hot) milling of a metal strip, e.g. steel strip in a steel producing factory.

The invention also relates to a method of steering a moving strip, in particular a metal strip, preferably a steel strip conveyed in a hot mill facility, in the machine direction, comprising a method of determining strip centre line position and orientation and/or camber of a moving strip, in particular a metal strip in a mill facility, according to the invention, as described above and if the centre line position and orientation and/or camber deviates from a set range, then the strip is displaced in a direction perpendicular to the machine direction, e.g. by feeding the obtained data in a control loop of a steering unit for lateral displacement of the moving strip.

According to another aspect of the invention a device is provided for determining a tracking characteristic selected from the group comprising strip centre line position and orientation, camber, and optionally strip width, of a hot moving strip, in particular a moving metal strip in a mill facility, preferably a hot steel strip in a hot mill, comprising a near infrared area scan camera for scanning a predetermined area covering at least the width of the strip and regions adjacent each side edge over a certain length of the moving strip, and a computer means for image processing of the images received from the area scan camera and for determining the edge positions of the strip and for calculating the tracking characteristic and optionally strip width from the edge positions thus determined. The device according to the invention basically comprises a near infrared area scan camera and a computer means which co-operate. The computer is programmed for image analysis and further calculations from the obtained data. The preferred embodiments discussed above with respect to the method(s) according to the invention are similarly applicable to this aspect of the invention and further aspects to be discussed. Advantageously the area scan camera is provided with a heat insulating element, such as a heat shield or (water) cooled housing.

Yet another aspect of the invention relates to a mill facility, in particular a hot mill facility, comprising a conveyor means for conveying the metal strip, preferably steel strip, in the machine direction, at least one roll station (mill stand) for reduction of the thiclcness of the metal strip, and a device for determining a characteristic value according to the invention. The advantages presented above regarding the methods according to the invention are similarly applicable to the device and mill facility according to the invention.

In a preferred embodiment the area scan camera is arranged such with respect to the metal strip to be analysed that the camera generates an image in perspective.

Advantageously, the mill facility comprises also an automated steering unit for lateral movement of the metal strip with respect to the conveyor means, wherein the automated steering unit is controlled by signals received from the computer means of the device for determining a characteristic value.

Advantageously the mill facility further comprises a marker positioned in the view area of the area scan camera as a reference for positioning the camera.

The invention is further illustrated by reference to the attached drawing, wherein

Fig. 1 shows diagrammatically a longitudinal view of an embodiment of a device for determining a characteristic value according to the invention, incorporated in a mill facility;

Fig. 2 shows diagrammatically a top view of the embodiment shown in Fig. 1 ; Fig. 3 shows a block diagram showing the several steps of the method according to the invention; and

Fig. 4 is flow diagram showing the basic steps of an embodiment of the data analysis used in the invention..

In Fig. 1 a mill facility is represented by two roll stations (mill stands) 10 and 12 respectively. The mill stands 10, 12 perform a thickness reduction of a moving strip 14 moving between a set of rolls 16, 18 in each mill stand 10, 12. A looper 20 for maintaining an appropriate tension of the moving strip 14 is provided in between the mill stand 10, 12. An area scan camera 22 (in a heat protective housing not separately shown) is mounted at the top of the mill stand 10 and is connected to computer 24 for computing data received from the camera 22. This computer 24 can also be part of a control loop for lateral movement means 26. The area scan camera 22 is mounted at an angle (e.g. 10-20°) with respect to the vertical and is directed opposite the machine direction. An alternative position would be on the mill stand 12 viewing in a downstream direction. The view area of the camera 22 is shown in Fig. 2. Due to the angled position of the camera, the view area 28 is in perspective. The computer 24 is designed for compensation of the perspective view as well as for compensating distortions associated with the camera lens. The view area is larger than the width of the strip and covers several meters of length of the strip. If desired, the exposure can be set to a desired value using additional illumination means not shown, also taking into account the temperature that is measured by a temperature sensor.

The operation of the device is schematically shown in the diagram of Fig. 3.

In a first step the position of the area scan camera is determined, and if necessary adjusted, e.g. between the processing of subsequent strips. Then the strip is conveyed through the mill stands in order to attain the desired thickness at the last mill stand. While the strip passes though the mill stands, the camera takes images at a sampling rate and view area and line speed of the strip such that subsequently taken images will overlap. The images taken are processed in an image analysing step, also using other input values such as strip thickness, temperature and looper position. From the contrast a continuous view of the edge lines of the strip is derived, which is less sensitive for the severe conditions that typically prevail in a hot mill. The thus derived edges are converted into a characteristic, which may be used if the value thereof is outside a predetermined range, for adjusting the lateral position of the strip. Optionally the position of the camera can be monitored during the milling process.

Fig. 4 shows a detailed diagram of an image processing embodiment. The raw data of an (distorted) image taken by the near infrared camera are loaded into the computer means. Based on contrast differences between strip and background both left and right edges are determined and only the data sets relating to both edges (edge pairs) are retained for further processing. These data sets are compensated for distortion resulting from the camera lens. Then the data sets thus compensated are corrected for the perspective view, thereby relating the pixel values to the actual position. Invalid data sets, i.e. outside predefined ranges, are removed. E.g. if the width should be in a window from 700 to 1200 mm, then a data set representing a width outside this window, e.g. 400 mm, is discarded. In a last step the (mean) centreline position, width and camber coefficients are calculated from the valid data sets.

Claims

Method of determining a tracking characteristic selected from the group comprising strip centre line position and orientation, camber, and optionally strip width, of a moving strip, in particular a moving metal strip in a mill facility, comprising an edge determination step using vision means, wherein repeatedly images of the moving strip are taken and the images are evaluated using image processing thereby determining the side edge positions of the strip, from which edge positions the tracking characteristic and optionally strip width is determined, wherein the vision means comprises a near infrared (NIR) area scan camera for scanning a predetermined area covering at least the width of the strip and regions adjacent each side edge over a certain length of the moving strip, and wherein the moving strip has a temperature of at least 800 °C.

Method according to claim 1 , wherein evaluation of the images comprises determining the strip side edges based on contrast between the strip and the regions adjacent each side edge.

Method according to claim 1 or claim 2, wherein the area scan camera is a CMOS video camera.

Method according to any one of the preceding claims, wherein during image processing distortion of the image due to the angle of view of the area scan camera with respect to the moving strip is compensated, after the side edge positions have been determined.

Method according to any one of the preceding claims, wherein during image processing distortion of the image due to the lens of the area scan camera is compensated, after the side edge positions have been determined.

6. Method according to any one of the preceding claims, wherein the position of the vision means with respect to the moving strip is watched at regular intervals.

7. Method according to any one of the preceding claims, wherein the sampling intervals of the vision means with respect to the moving strip are such that images subsequently taken in time are synchronized with strip position or strip speed of subsequently processed strips.

8. Method according to any one of the preceding claims, wherein the sampling intervals of the vision means with respect to the moving strip are such that images subsequently taken in time have some overlap or adjoin one another.

Method according to any one of the preceding claims, wherein the moving strip is a hot steel strip having a temperature in the range of 810-1200 °C, being treated in a hot mill facility.

Method of steering a moving strip, in particular a metal strip conveyed in a mill facility in the machine direction, comprising a method of determining strip centre line position and orientation, or camber of a moving strip according to any one of the preceding claims 1-9 and if the centre line position and orientation, or camber deviates from a set range, then a control loop is provided with data to correct the centre line position and/or orientation and/or camber.

A device for determining a tracking characteristic selected from the group comprising strip centre line position and orientation, camber and optionally strip width, of a moving strip, in particular a moving metal strip in a mill facility, comprising a near infrared area scan camera for scanning a predetermined area covering at least the width of the strip and regions adjacent each side edge over a certain length of the moving strip, and a computer means for image processing of the images received from the area scan camera and for determining the edge positions of the strip and for calculating the tracking characteristic and optionally strip width from the edge positions thus determined.

12. Device according to claim 11 , wherein the area scan camera is provided with a heat isolating element.

13. Device according to any one of the preceding claims 1 1 -12, wherein the area scan camera is a CMOS video camera.

14. Mill facility for rolling a metal strip comprising a conveyor means for conveying the metal strip in the machine direction, at least one roll station (mill stand) for reduction of the thickness of the metal strip, and a device according to any one of the previous claims 1 1-13.

15. Mill facility according to claim 14, further comprising an automated steering unit for lateral movement of the metal strip with respect to the conveyor means, the automated steering unit being controlled by signals received from the computer means of the device for determining a characteristic value.

16. Mill facility according to any one of claims 14 or 15, further comprising a marker positioned in the view area of the area scan camera as a reference point for positioning the camera.