JP2009192432A - Roll caliber position detector and roll caliber position detection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To precisely detect the relative position relationship of caliber parts in each caliber roll for adjusting the shape of a roll caliber. <P>SOLUTION: A control body section comprises: an illumination light source 7 for acquiring a sectional shape in a rolling direction, including the entire caliber of a roll caliber that is a rolling space formed by combining a plurality of caliber rolls, an optical fiber 8, a light projection lens 6; a photographic lens 9, an imaging section 10, and an image input section 12; an image processing section 13 and a roll identification section 14 for identifying each of the caliber rolls, based on the acquired sectional shape in a rolling direction; an operation section 15 for obtaining the shape profile of the caliber section of each of the identified caliber rolls, extracting each effective caliber curve contributing to rolling in each shape profile, and calculating a center position for drawing each effective caliber curve; and a display section 16 for calculating and outputting the relative positions of respective caliber rolls, based on respective calculated center positions. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a roll caliber position detection device and a roll caliber position detection method capable of accurately detecting the relative positional relationship of each caliber part of a roll caliber that is a rolling space formed by combining a plurality of caliber rolls. is there.

  Conventionally, a rolling space formed by combining a plurality of, particularly 3 to 4 or more grooved rolls (caliber rolls) is called a roll caliber, and various cross sections are obtained by passing a material to be rolled through the caliber roll and applying a reduction. A rolled material having a shape is manufactured. In order to increase the accuracy of the cross-sectional shape of the rolled material, it is important to cut or process each caliber roll with high accuracy and adjust the spatial arrangement of each caliber roll.

  As an adjustment method of the roll caliber, each adjuster adjusts the fitting state and the gap between the roll caliber and the reference rod by visually observing the fitting state and the gap between the roll caliber and the reference caliber processed to a predetermined outer diameter with high accuracy. The method of correcting the caliber roll arrangement, or using a projector to project the roll caliber onto the screen, the adjuster compares the projected roll caliber shape with the reference shape and corrects the arrangement of each caliber roll. There is a way to do it. However, any of the adjustment methods is a kind of sensory test depending on the visual observation of the adjuster, and there are problems in terms of reproducibility and objectivity, and there is a problem that time and labor for adjustment are large.

  In order to solve such a problem, in Patent Document 1, an image of a roll caliber edge is imaged using a rotating camera, an edge end position is detected by binarization processing, and the detected edge end position is detected. Based on the above, the shape and position of the roll caliber are precisely measured.

  Further, in Patent Document 2, a camera device that projects parallel light beams toward a roll caliber, enlarges a light pattern shielded by the roll caliber with a lens device, and projects the pattern onto a screen. Using the camera moving device that moves along the roll caliber, the contour shape of the entire roll caliber is acquired, and the adjustment amount of each caliber roll is calculated from the acquired contour data of the roll caliber. What corrects a roll caliber shape is described.

  However, in both Patent Documents 1 and 2, the camera is mechanically moved in order to obtain the entire shape of the roll caliber, so the mechanical error of the moving mechanism is superimposed on the measurement result, and the measurement accuracy deteriorates. End up. In general, the movement backlash of a slide mechanism such as a linear guide is about 20 μm or more. Therefore, in measuring the shape of the roll caliber, an error of about several tens mm to several mm occurs in the radius of curvature.

  On the other hand, in Patent Document 3, the imaging system is set to a magnification such that the entire roll caliber fits in one field of view, and an entire image of the roll caliber is created by one imaging, and each caliber roll is created from the created image. Extract the center contour, end contour, and taper contour, calculate the outermost dimension and contour center point from the center contour, and calculate and adjust the caliber roll corner points from the end contour and taper contour. The roll caliber shape is adjusted so that the corner point of the caliber roll is a predetermined position with respect to the corner point of the corresponding caliber roll. In this patent document 3, since the whole roll caliber is acquired by one imaging, the problem that the mechanical error by movement of imaging means, such as a camera, is superimposed on a measurement result is solved.

Japanese Patent No. 2595413 JP-A-8-5354 Japanese Patent No. 3835640

  By the way, in Patent Document 3, the roll caliber shape is adjusted based on the corner point of the caliber roll. The corner point of this caliber roll is a portion of the roll caliber that does not contribute to rolling, and the arc center of the roll caliber is not necessarily located at the center of both corner points of the caliber roll in the roll rotation axis direction. The precision control of grinding near the corner point is not high. Therefore, even if the roll caliber shape is adjusted based on the relative positional relationship between the corner points of the caliber roll, there is a problem that the arc portion of the roll caliber cannot often be accurately adjusted.

  Furthermore, in order to reliably detect the corner point of the caliber roll by the image, it is necessary that the silhouette of the tapered portion of the caliber roll is clearly imaged. Since it is a part where machine oil or the like adheres remarkably when it is incorporated into the stencil, the image tends to be unclear (see FIG. 3 of Patent Document 3), and the specific accuracy of the caliber roll corner points may deteriorate. There was a point. In order to solve this problem, it is conceivable to cut the entire caliber roll including the taper portion with high accuracy or to incorporate it into the housing so that machine oil does not adhere to the vicinity of the corner point. A new problem arises in that the cost and the built-in workload for the production and maintenance of caliber rolls increase.

  The present invention has been made in view of the above, and a roll caliber position detecting device and a roll caliber position capable of detecting the relative positional relationship of the caliber portion of each caliber roll for roll caliber shape adjustment with high accuracy. An object is to provide a detection method.

  In order to solve the above-described problems and achieve the object, a roll caliber position detection device according to the present invention includes a cross-sectional shape in the rolling direction including all calibers of a roll caliber that is a rolling space formed by combining a plurality of caliber rolls. A cross-sectional shape acquiring means for acquiring the shape, a roll identifying means for identifying each caliber roll based on the acquired cross-sectional shape in the rolling direction, and determining a shape profile of the caliber portion of each identified caliber roll, Extracts the effective caliber curves that contribute to rolling in, calculates the center position for calculating the center position for drawing each effective caliber curve, and calculates the relative position of each caliber roll based on each calculated center position And output means for outputting.

  Further, in the roll caliber position detection device according to the present invention, in the above invention, the center position calculation means includes a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve, An intersection point with the effective caliber curve is obtained, and a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance in a direction in which the shape profile is concave from the intersection is a center candidate point of the effective caliber curve And the process is repeated for two different sets of points to calculate a plurality of center candidate points, and the center position is determined based on the positions of the plurality of center candidate points. The processing is performed for each caliber roll.

  Further, in the roll caliber position detection device according to the present invention, in the above invention, the center position calculation means includes a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve, An intersection point with the effective caliber curve is obtained, and a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance in a direction in which the shape profile is concave from the intersection is a center candidate point of the effective caliber curve And calculating the plurality of center candidate points by calculating the variance value of the positions of the plurality of center candidate points, and calculating the variance value. The predetermined radius of curvature is reset so as to calculate the plurality of center candidate points such that the variance value is equal to or less than the minimum value or the predetermined value. Comprising the determined multiple central candidate points, and performing processing of determining the central position based on the position of the center candidate points plurality of each caliber roll.

  In the roll caliber position detection device according to the present invention as set forth in the invention described above, the center position calculation means determines a weighted average position of the plurality of center candidate points as the center position.

  In the roll caliber position detection device according to the present invention, in the above invention, the output means outputs a deviation between the center position of a caliber roll serving as a reference and the center position of another caliber roll. Features.

  Further, in the roll caliber position detection device according to the present invention, in the above invention, the output means includes the predetermined radius of curvature and the radius of curvature of each caliber roll whose variance value is a minimum value or a predetermined value or less. A deviation is output.

  Also, the roll caliber position detection method according to the present invention detects the relative position of each caliber roll by acquiring a rolling direction cross-sectional shape including all calibers of the roll caliber that is a rolling space formed by combining a plurality of caliber rolls. In the roll caliber position detection method, the effective caliber curve that identifies each caliber roll based on the acquired cross-sectional shape in the rolling direction, determines the shape profile of the caliber part of each caliber roll, and contributes to rolling within each shape profile A center position calculating step for calculating a center position for drawing each effective caliber curve, and a relative position output step for calculating and outputting a relative position of each caliber roll based on each center position. It is characterized by including.

  In the roll caliber position detection method according to the present invention, in the above invention, the center position calculation step may include a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve, An intersection point with the effective caliber curve is obtained, and a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance in a direction in which the shape profile is concave from the intersection is a center candidate point of the effective caliber curve And a center candidate point calculating step of calculating a plurality of center candidate points by repeatedly performing the process for other two different sets of points, and based on the positions of the plurality of center candidate points The center position determining step for determining the center position is performed for each caliber roll, and the center position for each caliber roll is calculated.

  In the roll caliber position detection method according to the present invention, in the above invention, the center position calculation step may include a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve, An intersection point with the effective caliber curve is obtained, and a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance in a direction in which the shape profile is concave from the intersection is a center candidate point of the effective caliber curve A center candidate point calculating step of calculating a plurality of center candidate points by repeatedly performing the process for two different sets of points and calculating a variance value of the positions of the plurality of center candidate points. The predetermined radius of curvature is reset so that the variance value is a minimum value or less than a predetermined value, and the processing of the center candidate point calculation step is repeated to obtain the variance value A variance value calculating step for obtaining the plurality of center candidate points that are a small value or less than a predetermined value, and a center position for determining the center position based on the positions of the plurality of center candidate points obtained by the variance value calculating step The determination step is performed for each caliber roll, and the center position for each caliber roll is calculated.

  In the roll caliber position detection apparatus and the roll caliber position detection method according to the present invention, the cross-sectional shape in the rolling direction including all the calibers of the roll caliber, which is a rolling space formed by combining a plurality of caliber rolls, is acquired. Each caliber roll is identified based on the cross-sectional shape in the rolling direction, the shape profile of the caliber part of each caliber roll is obtained, the effective caliber curve contributing to rolling is extracted from each shape profile, and each effective caliber curve is drawn. Center position is calculated, and the relative position of each caliber roll is calculated and output based on each center position, so the relative position of the caliber part of each caliber roll for roll caliber shape adjustment The relationship can be detected with high accuracy.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of a roll caliber position detection device and a roll caliber position detection method according to the present invention will be described with reference to the accompanying drawings. Note that the present invention is not limited to the embodiments. In the description of the drawings, the same or corresponding parts are denoted by the same reference numerals.

(Embodiment 1)
First, a first embodiment of the present invention will be described. FIG. 1 is a schematic diagram showing a schematic configuration of a roll caliber position detection device according to Embodiment 1 of the present invention. In addition, in this Embodiment 1, the roll caliber position detection apparatus at the time of applying to the 4 roll mill housing which is a kind of caliber roll mill is demonstrated.

  In FIG. 1, this roll caliber position detection apparatus has a roll 2 to be measured, which is composed of a plurality of caliber rolls, in a bearing housing 1, and the relative positions of the caliber rolls for adjusting the roll caliber shape of the roll 2. The position is detected.

  In the bearing housing 1, as shown in FIG. 2, four caliber rolls 2 a to 2 d are disposed so as to surround the rolled material 22. That is, the rolls 2a and 2b are arranged to face each other in the X direction, and the rolls 2c and 2d are arranged to face each other in the Y direction. However, the rolled material 22 is removed when the roll caliber position is detected. Each of the rolls 2a to 2d has rolling shafts 21a to 21d that are rotation axes of the respective rolls 2a to 2d, and the rolling direction of each of the rolls 2a to 2d and the axial direction of the rolling shafts 21a to 21d are arranged orthogonal to each other. The Therefore, when adjusting the roll caliber, each of the rolls 2a to 2d needs to adjust a direction parallel to the reduction direction (a reduction direction) and a direction orthogonal to the reduction direction (a thrust direction). For example, the roll 2a needs to adjust the X direction that is the reduction direction and the Y direction that is the thrust direction.

  In order to detect the position of the roll caliber of the roll 2 in the bearing housing 1, the roll caliber position detection device shown in FIG. 1 includes an illumination light source 7 that emits illumination light, an optical fiber 8 that guides the illumination light, From the light projection lens 6 that converts the illumination light guided from the optical fiber 8 into parallel light of a substantially cylindrical shape that covers the roll caliber as in the region E shown in FIG. 2 and is orthogonal to the rolling axes 21a to 21d. It has a floodlight system. In FIG. 1, the illumination light source 7 and the projection lens 6 are separated via the optical fiber 8, but the optical fiber 8 is deleted and the illumination light source 7 and the projection lens 6 are integrated. You may comprise.

  In addition, the roll caliber position detection device is an imaging lens that forms on the imaging surface a bright and dark image as shown in FIG. 3 that is a part of the parallel light that has passed through the roll caliber and is generated by blocking the caliber portion of the roll 2. 9 and an imaging unit 10 that captures the bright and dark image input to the imaging surface described above, and a connection cable 10a that transmits the image data captured by the imaging unit 10 to the control main body unit 11. The imaging lens 9 is a lens that forms an object-side telecentric optical system, and the optical axis C of this optical system is preferably parallel to the pass line of the material to be rolled and coincides with the optical axis of the light projecting system. . The imaging unit 10 is realized using an area sensor such as a two-dimensional CCD. However, when it is necessary to increase the image resolution, the imaging unit 10 is realized using an optical axis sweeping means using a galvanometer mirror and a line sensor. May be.

  Here, assuming that the maximum rolling diameter by roll 2 is 90 mm and the roll caliber adjustment limit is 0.1 mm, the pixel resolution is 90/2048 = 0.044 mm by using an area sensor of 2048 × 2048 pixels. Sufficient resolution can be obtained.

  The control main body 11 includes a roll caliber as shown in FIG. 4 from the image input unit 12 as an input interface for receiving image data input via the connection cable 11a and the image data input to the image input unit 12. An image processing unit 13 that generates an edge image, a roll identification unit 14 that identifies each of the four rolls 2a to 2d from the edge image, and each roll 2a to 2d that is identified by the roll identification unit 14 are shown in FIG. A profile PF in which only the caliber portion is cut out is generated, an effective caliber curve of an effective caliber region RE contributing to rolling is extracted from the profile PF, and a center position for calculating the center position for drawing this effective caliber curve is calculated. The calculation unit 15 as a calculation means and the relative positions of the rolls 2a to 2d based on the center positions of the rolls 2a to 2d. The display unit 16 serving as an output means for displaying and outputting the position is realized by a CPU and the like, and at least controls the image input unit 12, the image processing unit 13, the roll identification unit 14, the calculation unit 15, the display unit 16, and the illumination light source 7. And a storage unit 18 that retains various programs and data and secures a recording area necessary for calculation. The light projecting system, the light receiving system, and the image input unit 12 function as a cross-sectional shape acquisition unit.

  Here, when calculating the center position, as shown in FIG. 6, the calculation unit 15 uses a vertical bisector of a line segment formed by a pair of different two points P1a and P1b on the effective caliber curve and the effective bisector. An intersection PC1 with the caliber curve is obtained, and a point having a distance R equal to a predetermined curvature radius (nominal caliber curvature radius) R of an effective caliber curve determined in advance in the direction in which the profile PF is concave from the intersection PC1 is effective. It is calculated as the center candidate point P1 of the caliber curve. Similarly, the intersection PC2 between the vertical bisector formed by the pair of the two points P2a and P2b and the effective caliber curve is obtained, and the nominal caliber curvature is oriented from the intersection PC2 in the direction in which the profile PF becomes concave. A point having a distance R equal to the radius R is calculated as a center candidate point P2. The same process is repeated for other two different sets of points to calculate a center candidate point group P consisting of a plurality of center candidate points, and the weighted average position of this center candidate point group P is the center for drawing the effective caliber curve. Find as position. That is, the calculating part 15 calculates | requires a center position also with respect to the profile corresponding to other three rolls. The display unit 16 displays and outputs the relative positional relationship between the rolls 2a to 2d with these four center positions as the positions of the rolls 2a to 2d.

  Note that widening jigs 3a and 3b are provided between the light projecting lens 6 and the imaging lens 9, and the roll gap is constrained so that the roll gap does not become unstable. That is, the widening jig 3a is provided between the light projecting lens 6 and the rolls 2a and 2b, and is inserted between the rolls 2a and 2b from the light projecting lens 6 side to widen the width between the rolls 2a and 2b. The widening jig 3b is provided between the imaging lens 9 and the rolls 2c and 2d, and is inserted between the rolls 2c and 2d from the imaging lens 9 side to widen the rolls 2c and 2d. The widening jigs 3a and 3b are formed in a cylindrical shape so that parallel light can pass therethrough. However, the widening jigs 3a and 3b are projected so as not to cause a ghost of reflected light from the inner surface. It is preferable to adjust the aperture of the optical lens 6, the inner diameters of the widening jigs 3a and 3b, and the numerical aperture of the light projection lens 6.

  As described above, it is preferable that the optical axis of the projection lens 6 and the optical axis of the imaging lens 9 coincide with each other and the pass lines of the bearing housing 1 are parallel or coincide with each other. However, in order to easily realize this positional relationship. In the first embodiment, the housing frame 4 is provided, and the bearing housing is fixed by arranging a stopper 5 on the upper surface of the housing frame 4 so as to surround the lower part of the bearing housing 1 from three directions. The stopper 5 is made to stop by a push bolt (not shown) so that the stopper 5 can be arranged and fixed easily and with good reproducibility. Further, the housing base 4 is provided with optical system holding mechanisms 19a and 19b for holding the light projecting lens 6 and the imaging lens 9, respectively. By the optical support mechanism 19a, 19b having the housing pedestal 4, the stopper 5, and the position adjusting mechanism, the above-described adjustment of the optical axis and the pass line is facilitated. In addition, it is preferable to provide a retracting mechanism that can retract the optical system holding mechanisms 19a and 19b to prevent damage when the bearing housing 1 is installed on the housing base 4.

  Here, the roll caliber position detection processing procedure according to the first embodiment will be described with reference to the flowchart shown in FIG. In FIG. 7, the control unit 17 causes the illumination light source 7 to output illumination light, causes the imaging unit 10 to capture a bright and dark image on which the roll caliber is projected, and acquires image data indicating the bright and dark image using the image input unit 12 ( Step S101). Thereafter, the image processing unit 13 generates an edge image including the caliber portion shown in FIG. 4 from the image data (step S102). Thereafter, the roll identifying unit 14 identifies each roll 2a to 2d from the edge image, and the calculation unit 15 cuts out four profiles PF of the caliber portion of each roll 2a to 2d (step S103).

  Furthermore, the calculating part 15 reads the nominal caliber curvature radius R preset to the roll caliber (step S104). Then, one profile PF is selectively read (step S105), the effective caliber area RE is set (step S106), and the variable n is set to the initial value 0 (step S107).

  Then, the computing unit 15 selects two points from the profile PF in the effective caliber region RE (step S108), calculates a perpendicular bisector connecting the two points (step S109), and calculates the vertical 2 The intersection of the bisector and the profile PF is obtained, and a point on the vertical bisector on the concave side of the profile PF and having a distance R from the intersection is set as a candidate point (center candidate point) of the caliber circle center. (Step S110).

  Thereafter, the variable n is incremented (step S111), and it is further determined whether or not the variable n exceeds a predetermined integer value N (step S112). When the variable n does not exceed the predetermined integer value N (No in step S112), the process proceeds to step S108, and the center candidate point is calculated until the variable n exceeds the predetermined integer value N.

  On the other hand, when the variable n exceeds the predetermined integer value N (step S112, Yes), the weighted average value of the obtained coordinate group composed of a plurality of center candidate points is set to the caliber center (center of the currently processed roll). Position) (step S113).

  Thereafter, it is determined whether or not there is a next profile PF (step S114). When there is the next profile PF (step S114, Yes), the process proceeds to step S105, the next profile PF is selectively read and the above-described processing is performed, and the caliber center position is determined. On the other hand, when there is no next profile PF (step S114, No), the display part 16 displays and outputs the deviation of the caliber center of another roll with respect to the caliber center of a reference roll (step S115), and complete | finishes this process.

  FIG. 8 is a diagram illustrating a display output example of the caliber center deviation by the display unit 16. As shown in FIG. 8, the roll 2a located on the left in FIG. 2 is used as a reference roll, and the roll 2c located on the top, the roll 2b located on the right, the roll 2b located on the right, and the roll 2d located on the bottom. Thus, the deviation ΔX with respect to the X direction and the deviation ΔY with respect to the Y direction are displayed and output. These deviations ΔX and ΔY are obtained based on the deviations between the caliber centers.

  If it is known that the nominal caliber curvature radius R and the radius φ of the material to be rolled are not equal, the deviation value including the correction value may be displayed and output as the correction amount of the roll caliber. In addition, roll caliber adjustment in an actual roll chock often cannot be moved by accurately setting the movement amount of the roll, so it is determined whether or not the deviation of each roll is within a predetermined allowable range. The determination result and the correction direction corresponding to the direction of deviation, for example, plus or minus in the reduction direction, plus or minus in the thrust direction may be displayed and output.

  FIG. 9 is a diagram showing the relationship between the movement amount of the calculation center of the roll caliber by the roll caliber position detection device shown in FIG. 1 and the actual measurement movement amount of the roll caliber. In FIG. 9, the center position of the caliber is calculated by the roll caliber position detection method according to the first embodiment while moving the roll in the down direction in the chock, and the amount of movement is actually measured with a high-precision laser distance meter. Results are shown. As shown in FIG. 9, in the reduction direction, the movement amount of the calculation center and the actual movement amount are almost the same and agree well, and the thrust direction is almost the same and there is almost no change. From this result, it was confirmed that the position can be detected with high accuracy by the position detection method of the roll caliber according to the first embodiment.

  In the first embodiment, an edge image is obtained from image data obtained by imaging the entire roll caliber, each roll is identified from the edge image, a center position for drawing the effective caliber curve of each identified roll is obtained, and the obtained center Since the position deviation is displayed and output as the deviation of each roll position, the position of the roll caliber can be detected with high accuracy, and the roll caliber adjustment can be performed based on the position detection result.

  In particular, in the first embodiment, since the entire roll caliber is imaged at a time, there is no movement of the imaging device as compared with the case where the caliber portion of each roll is moved and imaged. Can be eliminated. In addition, the center position for drawing the effective caliber curve is obtained only from the effective caliber curve of the caliber part without using the corner point of the caliber roll as a reference, and the relative positional relationship of each caliber roll is determined by this center position. Therefore, it is possible to always detect the position of the roll caliber with high accuracy. In addition, a problem that occurs when highly accurate position detection is performed with reference to the corner point of the caliber roll, that is, the manufacture and maintenance of the caliber roll that cuts the entire caliber roll including the tapered portion with high precision. The position of the caliber roll can be detected simply and with high accuracy without increasing the cost and increasing the work load of incorporating into the housing so that machine oil or the like does not adhere to the corners.

(Embodiment 2)
Next, a second embodiment of the present invention will be described. In the first embodiment described above, the nominal caliber radius of curvature R is used for calculation of the center point of the caliber and the calculation of the center. However, in this second embodiment, the radius of curvature of the caliber can be calculated simultaneously. Yes.

  10 to 14 are diagrams showing the results of calculating the caliber center candidate point using different radii of curvature for the same profile. 10 to 14, the curvature radii of the profiles are all 993.040 (pixels), and the curvature radii (set curvature radii) used for calculating the center candidate point are 214.800 → 603.920 → 993, respectively. .040 → 1402.640 → 19995.560 (pixels). From the results of FIGS. 10 to 14, when the curvature radius of the profile is substantially equal to the set curvature radius, the two-dimensional dispersion (variation) of the center candidate point is small, and the center is set as the set curvature radius deviates from the curvature radius of the profile. It can be seen that the variation of candidate points increases.

  FIG. 15 is a diagram illustrating the dependence of the variation of the center candidate point on the set radius of curvature. As shown in FIG. 15, the variation of the center candidate point is a simple concave concave curve with respect to the set radius of curvature, and there is a set radius of curvature that minimizes the variation of the center candidate point. It can be said that the set radius of curvature is approximately equal to the radius of curvature of Caliber. Therefore, if the variation of the center candidate point is obtained based on the set radius of curvature set as an appropriate initial value and is reset to the set radius of curvature that minimizes this variation, the minimum value will eventually be obtained. The set radius of curvature can be calculated as the radius of curvature of the caliber. This setting radius resetting process can be realized by using a minimum value search algorithm such as a local search method including a hill-climbing method. The configuration of the roll caliber position detection device according to the second embodiment is the same as that of the first embodiment shown in FIG. 1, but the processing by the calculation unit 15 is different.

  Here, with reference to the flowchart shown in FIG. 16, the roll caliber position detection processing procedure including the calculation process of the curvature radius of the caliber according to the second embodiment will be described. In FIG. 16, as in the first embodiment, the control unit 17 causes the illumination light source 7 to output illumination light, causes the image capturing unit 10 to capture a bright and dark image, and captures image data indicating the light and dark image. Obtained by the image input unit 12 (step S201). Thereafter, the image processing unit 13 generates an edge image including a caliber portion from the image data (step S202). Thereafter, the roll identifying unit 14 identifies each roll 2a to 2d from the edge image, and the calculation unit 15 cuts out four profiles PF of the caliber portion of each roll 2a to 2d (step S203).

  Further, the calculation unit 15 reads the nominal caliber curvature radius R and sets it as an initial value of the set curvature radius (step S204). Then, one profile PF is selectively read (step S205), the effective caliber area RE is set (step S206), and the variable n is set to the initial value 0 (step S207).

  Then, the computing unit 15 selects two points from the profile PF in the effective caliber region RE (step S208), calculates a perpendicular bisector connecting the two points (step S209), and calculates the vertical 2 The intersection of the bisector and the profile PF is obtained, and a point on the vertical bisector on the concave side of the profile PF and having a distance R from the intersection is set as a candidate point (center candidate point) of the caliber circle center. (Step S210).

  Thereafter, the variable n is incremented (step S211), and it is further determined whether or not the variable n exceeds a predetermined integer value N (step S212). When the variable n does not exceed the predetermined integer value N (No in step S212), the process proceeds to step S208, and the center candidate point is calculated until the variable n exceeds the predetermined integer value N.

  On the other hand, when the variable n exceeds the predetermined integer value N (step S212, Yes), the variation E (R) of the coordinate group including the obtained plurality of center candidate points is calculated (step S213). Then, it is determined whether or not the variation E (R) is a minimum value using a minimum value search algorithm (step S214). If the variation E (R) is not the minimum value (No in step S214), the changed setting is made by changing the set radius of curvature so that the variation E (R) becomes the minimum value based on the minimum value search algorithm. The radius of curvature is reset and the process proceeds to step S207 (step S215), and the process of obtaining a plurality of center candidate points using the reset radius of curvature is repeated.

  On the other hand, when the variation E (R) is the minimum value (step S214, Yes), the weighted average position of the coordinate group including the plurality of center candidate points at which the variation E (R) has the minimum value is obtained. While determining as the center position of the roll caliber, the currently set curvature radius is determined as the curvature radius of the roll caliber (step S216).

  Thereafter, similarly to the first embodiment, it is determined whether or not there is a next profile PF (step S217). When there is the next profile PF (step S217, Yes), the process proceeds to step S205, the next profile PF is selectively read and the above-described processing is performed to determine the center position of the roll caliber and the curvature of the roll caliber. Determine the radius. On the other hand, when there is no next profile PF (step S217, No), the display unit 16 displays and outputs the deviation of the other caliber center with respect to the caliber center of the reference roll and the deviation of the curvature radius of each caliber with respect to the nominal caliber curvature radius. (Step S218), this process is terminated.

  FIG. 17 is a diagram illustrating an example of display output by the display unit 16. As shown in FIG. 17, in addition to the display contents shown in FIG. 7, the deviation ΔR of the curvature radius of each roll 2 a to 2 d with respect to the nominal caliber curvature radius R is displayed and output as a percentage. Based on the display output contents shown in FIG. 17, it is possible to know not only the relative positional relationship of each caliber roll but also the grinding state of the caliber of each caliber roll. For example, the left and right rolls 2a and 2b are cut to be larger than the nominal caliber curvature radius R which is a grinding set value, and the upper and lower rolls 2c and 2d are cut to be smaller than the nominal caliber curvature radius R. As a result, it is possible to manage the tendency of the grinding accuracy of the roll caliber.

  In the processing procedure shown in FIG. 16, the variation E (R) of a plurality of center candidate points is first obtained. If the variation E (R) is not the minimum value, the caliber obtained from the plurality of center candidate points is obtained. Although the center position is not calculated, the present invention is not limited to this, and the center position of the caliber may be calculated every time a plurality of center candidate points are obtained. In step S214, the radius of curvature is determined only when the variation E (R) is the minimum value. However, the present invention is not limited to this. In step S214, the variation E (R) is less than or equal to the predetermined value. In this case, the center position and the radius of curvature of the caliber may be determined. In this case, if it is within an allowable error range, the center position and the radius of curvature of the caliber can be quickly determined.

  In the second embodiment, the nominal caliber radius of curvature R is used as a set radius of curvature as an initial value. However, the present invention is not limited to this, and a set radius of curvature within another appropriate predetermined range may be used as an initial value. .

  In the second embodiment, not only the center position of the caliber but also the radius of curvature of the caliber can be determined, so that the relative positional relationship between the caliber rolls can be detected with high accuracy and the grinding accuracy of the caliber can be improved. Trend management can also be performed.

  In the first and second embodiments described above, when measuring the shape of the roll caliber, the light projecting system projects parallel light, and the light receiving system uses a telecentric optical system to obtain a roll caliber silhouette image. However, the present invention is not limited to this, and for example, the shape of the roll caliber may be measured using a method by moving a touch probe type or triangulation laser type distance sensor. In short, it is only necessary to be able to measure the shape of the roll caliber, and the center for drawing the effective caliber curve may be obtained based on the shape of the roll caliber.

Further, in the first embodiment described above, the center position of the caliber is obtained using the nominal caliber radius of curvature R, and in the second embodiment described above, the center position of the caliber is determined using the set curvature radius having the minimum variation. However, it is preferable that the processing according to the first and second embodiments can be appropriately switched.

  Furthermore, in Embodiments 1 and 2 described above, a vertical bisector is obtained from the effective caliber curve, and the nominal caliber radius of curvature or variation is minimized from the intersection of the vertical bisector and the effective caliber curve. The position of the point up to the distance of the set radius of curvature is determined as the center candidate point. However, the present invention is not limited to this, and for example, the intersection position of the perpendicular bisectors may be determined as the center candidate point. .

It is a schematic diagram which shows schematic structure of the roll caliber position detection apparatus which is Embodiment 1 of this invention. It is a figure which shows schematic structure in the bearing housing shown in FIG. It is a figure which shows an example of the bright and dark image which an imaging part receives. It is a figure which shows an example of the edge image which an image process part produces | generates. It is a figure which shows an example of the profile which cut out only the caliber part. It is the figure which showed typically the calculation method of a center candidate point. It is a flowchart which shows the roll caliber position detection processing procedure by Embodiment 1 of this invention. It is a figure which shows an example of the display output by a display part. It is a figure which shows the relationship between the movement amount of the calculation center of a roll caliber by the roll caliber position detection apparatus shown in FIG. 1, and the actual measurement movement amount of a roll caliber. It is a figure which shows an example of the result of having calculated the center candidate point of the caliber using a different curvature radius with respect to the same profile. It is a figure which shows an example of the result of having calculated the center candidate point of the caliber using a different curvature radius with respect to the same profile. It is a figure which shows an example of the result of having calculated the center candidate point of the caliber using a different curvature radius with respect to the same profile. It is a figure which shows an example of the result of having calculated the center candidate point of the caliber using a different curvature radius with respect to the same profile. It is a figure which shows an example of the result of having calculated the center candidate point of the caliber using a different curvature radius with respect to the same profile. It is a figure which shows the setting curvature radius dependence of the dispersion | variation in a center candidate point. It is a flowchart which shows the roll caliber position detection process procedure including the calculation process of the curvature radius of the caliber by Embodiment 2 of this invention. It is a figure which shows an example of the display output by the display part by Embodiment 2 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Bearing housing 2, 2a-2d Roll 3a, 3b Widening jig | tool 4 Housing mount 5 Stopper 6 Light projection lens 7 Illumination light source 8 Optical fiber 9 Imaging lens 10 Imaging part 10a Connection cable 11 Control part main body 12 Image input part 13 Image processing Part 14 Roll identification part 15 Calculation part 16 Display part 17 Control part 18 Storage part 21a-21d Rolling axis

Claims (9)

Cross-sectional shape acquisition means for acquiring a rolling direction cross-sectional shape including all calibers of a roll caliber that is a rolling space formed by combining a plurality of caliber rolls;
Roll identifying means for identifying each caliber roll based on the acquired rolling direction cross-sectional shape;
A center position calculating means for obtaining a shape profile of a caliber portion of each identified caliber roll, extracting an effective caliber curve contributing to rolling in each shape profile, and calculating a center position for drawing each effective caliber curve; ,
Output means for calculating and outputting the relative position of each caliber roll based on each calculated center position;
A roll caliber position detecting device comprising:   The center position calculating means obtains an intersection point between a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve and the effective caliber curve, and the shape profile is concave from the intersection point. And calculating a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance as a center candidate point of the effective caliber curve, and converting the process into another set of two points. 2. The method according to claim 1, wherein a plurality of center candidate points are calculated by repeatedly performing the process, and the center position is determined for each caliber roll based on the positions of the plurality of center candidate points. Roll caliber position detection device.   The center position calculating means obtains an intersection point between a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve and the effective caliber curve, and the shape profile is concave from the intersection point. And calculating a point having a distance equal to a predetermined radius of curvature of the effective caliber curve determined in advance as a center candidate point of the effective caliber curve, and converting the process into another set of two points. It is repeatedly performed to calculate a plurality of center candidate points, and further, a variance value of the positions of the plurality of center candidate points is obtained, and the predetermined curvature radius is reset so that the variance value becomes a minimum value or a predetermined value or less. Repeatedly calculating the plurality of center candidate points to obtain the plurality of center candidate points having a variance value that is a minimum value or a predetermined value or less, and based on the positions of the plurality of center candidate points, Roll caliber position detecting device according to the processing to determine the position to claim 1, characterized in that for each caliber roll.   The roll caliber position detection apparatus according to any one of claims 1 to 3, wherein the center position calculation unit determines a weighted average position of the plurality of center candidate points as the center position.   The roll caliber according to any one of claims 1 to 4, wherein the output means outputs a deviation between the center position of a caliber roll serving as a reference and the center position of another caliber roll. Position detection device.   The output means outputs a deviation between the predetermined radius of curvature and the radius of curvature of each caliber roll whose variance value is a minimum value or less than a predetermined value. Roll caliber position detection device as described in 1. In the roll caliber position detection method for detecting the relative position of each caliber roll by acquiring a rolling direction cross-sectional shape including all calibers of the roll caliber that is a rolling space formed by combining a plurality of caliber rolls,
Each caliber roll is identified based on the acquired cross-sectional shape in the rolling direction, the shape profile of the caliber part of each caliber roll is obtained, and the effective caliber curve that contributes to rolling is extracted within each shape profile. A center position calculating step for calculating a center position for drawing a curve;
Relative position output step for calculating and outputting the relative position of each caliber roll based on each center position;
A roll caliber position detection method comprising: The center position calculating step includes:
An intersection point between a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve and the effective caliber curve is obtained, and the shape profile is determined in advance from the intersection so that the shape profile is concave. A point having a distance equal to a predetermined radius of curvature of the effective caliber curve is calculated as a center candidate point of the effective caliber curve, and the process is repeated for another set of two different points. A center candidate point calculating step for calculating a center candidate point;
A center position determining step for determining the center position based on the positions of the plurality of center candidate points;
8. The roll caliber position detection method according to claim 7, wherein the center position for each caliber roll is calculated for each caliber roll. The center position calculating step includes:
An intersection point between a vertical bisector of a line segment formed by a set of two different points on the effective caliber curve and the effective caliber curve is obtained, and the shape profile is determined in advance from the intersection so that the shape profile is concave. A point having a distance equal to a predetermined radius of curvature of the effective caliber curve is calculated as a center candidate point of the effective caliber curve, and the process is repeated for another set of two different points. A center candidate point calculating step for calculating a center candidate point;
A variance value of the positions of the plurality of center candidate points is obtained, the predetermined radius of curvature is reset so that the variance value becomes a minimum value or a predetermined value or less, and the processing of the center candidate point calculation step is repeated to perform the variance A variance value calculating step for obtaining the plurality of center candidate points whose value is a minimum value or a predetermined value or less;
A center position determining step for determining the center position based on the positions of the plurality of center candidate points obtained by the variance value calculating step;
8. The roll caliber position detection method according to claim 7, wherein the center position for each caliber roll is calculated for each caliber roll.

Images