Method for measuring the shape of a material to be measured, apparatus for measuring the shape of a material to be measured, and method for manufacturing a product.

Abstract

To provide a technique capable of automatically, accurately, and efficiently measuring a shape of a material to be measured.SOLUTION: A shape measurement method for measuring a shape of a material 1 to be measured includes: an area specifying step of acquiring two or more feature point existing areas SA, each of which is an area in which a feature point for specifying the shape of the material 1 to be measured is estimated to exist; a precise measurement step of acquiring a surface shape of specified each feature point existing area SA as precise measurement data composed of a set of three-dimensional point groups; and a feature point position calculation step of obtaining information on a position P of the feature point existing in each feature point existing area SA from the acquired precise measurement data.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a technique for more accurately measuring the shape of a measurement target material such as a metal plate, and a technique for manufacturing a product using this technique. The present invention is a particularly effective technique for a measurement target material whose reference shape when viewed from the measurement direction is known in advance.

Background Art

[0002] For example, in the thick plate refining process of a steel mill, there is a cutting process for cutting a rolled material rolled in a rolling process to a specified dimension. As the cutting method, for example, a cutting method such as gas cutting, laser cutting, or plasma cutting is adopted. The work itself of the cutting process has advanced in automation by using a numerically controlled cutting device or a self-propelled cutting device with a portal-shaped movable part as a cutting device.

[0003] However, the dimensional measurement and flatness measurement of the thick plate (measurement target material) before and after cutting are still mostly manual operations. Such measurement operations such as the shape measurement of the thick plate after cutting by hand are heavy labor operations. In addition, the dimensional measurement of the thick plate is an operation in which there is a risk of falling disasters due to poor working platforms and entry work into the production line, and there is a risk of collisions and narrow pressure due to interference between workers and the cutting device. Also, due to manual measurement, there are issues regarding accuracy, including concerns about individual differences and human errors, and the inability to take many measurement points.

[0004] On the other hand, as a method for measuring the dimensions of a thick plate without relying on manual labor, a non-contact measurement method using optical means has been disclosed. For example, Patent Document 1 discloses a measurement method using a sensor system that uses a light cutting method to measure the shape of a measurement target material by irradiating the measurement target material with laser light and imaging the irradiated laser from another angle with a camera. Also, Patent Document 2 discloses mounting a non-contact distance sensor on a portal-shaped cutting device and calculating the outer dimensions of the cut material (measurement target material) from the sensing data of the distance sensor.

Prior Art Documents

[0005] [Patent Document 1] Patent No. 6780533 [Patent Document 2] Japanese Patent Publication No. 2000-158169 [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] The method described in Patent Document 1 uses the light section method to measure the shape of the material to be measured. Therefore, it is possible to measure the shape of the material to be measured with high accuracy. However, generally, the width dimension of thick plates cut in the thick plate refining process is large, ranging from about 1000 mm to 5000 mm, and the placement position of the thick plate to be measured is not constant. On the other hand, if one attempts to accurately measure the dimensions of a thick plate with a resolution of 1 mm or less using a sensor system based on the light section method, it is only possible to measure a range of about 400 mm in width at a time. Thus, when attempting to measure thick plates with large widths, where the placement position of the thick plates is not constant, using a sensor system based on the light section method, it is necessary to set up multiple sensors. For this reason, the method described in Patent Document 1 is costly and inefficient.

[0007] Furthermore, in the cutting apparatus described in Patent Document 2, when measuring the shape data of the material to be cut, sensing is performed by scanning the distance sensor in a zigzag pattern across the entire area of ​​the material to be cut. As a result, the method in Patent Document 2 has an increased cycle time and an enormous amount of measurement data, resulting in poor efficiency.

[0008] This invention was made with the above-mentioned points in mind, and one of its objectives is to provide a technology that can automatically, accurately, and efficiently measure the shape of a material to be measured. [Means for solving the problem]

[0009] The inventor conceived of a method to obtain rough information on the position of each feature point that identifies the shape of the material being measured using a simple method, and then accurately measure the area (narrow area) of this rough information to accurately obtain the position of each feature point that can identify the shape of the material being measured. The inventor found that this method allows for simple and accurate shape measurement even when the dimensions of the material being measured are large. Based on this finding, the inventor made the present invention.

[0010] To solve the problem, one aspect of the present invention is a shape measurement method for measuring the shape of a material to be measured, comprising: a region identification step of acquiring two or more feature point existence regions, which are regions where it is estimated that feature points that identify the shape of the material to be measured exist; a precision measurement step of acquiring the surface shape of each identified feature point existence region as precision measurement data consisting of a set of three-dimensional point clouds; and a feature point position calculation step of determining the position information of the feature points present in each feature point existence region from the acquired precision measurement data.

[0011] Another aspect of the present invention is a shape measuring device for measuring the shape of a material to be measured, comprising: a region identification unit that acquires two or more feature point existence regions, which are regions where it is estimated that feature points that identify the shape of the material to be measured exist; a precision measurement unit that acquires the surface shape of each acquired feature point existence region as precision measurement data consisting of a set of three-dimensional point clouds; and a feature point position calculation unit that calculates the position information of the feature points located in the corresponding feature point existence region from the acquired precision measurement data. [Effects of the Invention]

[0012] According to an aspect of the present invention, for example, the shape of the material to be measured is roughly measured to extract the region where feature points exist. Then, by precisely measuring the surface shape of the narrowed region where feature points exist, positional information of two or more feature points that can identify the shape of the material to be measured is obtained. As a result, according to an embodiment of the present invention, the shape of the material to be measured can be measured automatically, accurately, and efficiently. Shape measurement includes, for example, dimensional measurement and profile measurement of the material to be measured. [Brief explanation of the drawing]

[0013] [Figure 1] This is a diagram showing examples of the positions of feature points and the regions where feature points exist. [Figure 2] This is a diagram showing an example of the processing configuration of a shape measurement method according to an embodiment based on the present invention. [Figure 3] This is an enlarged view showing an example of the relationship between the position P′ of the feature point obtained by rough measurement, the feature point existence region SA based on the position P′ of the feature point, and the position P of the actual feature point. [Figure 4] This is a schematic diagram showing the configuration of a shape measurement device according to an embodiment based on the present invention. (a) is a side view and (b) is a plan view. [Figure 5] This is a diagram showing an example of the support state of the measurement target material placed in the cutting region. [Figure 6] This is a diagram showing the configuration of the shape calculation unit. [Figure 7] This is a schematic diagram for explaining rough measurement. (a) is a side view and (b) is a plan view. [Figure 8] This is a diagram showing an example of the processing flow of the extraction unit. [Figure 9] This is a diagram for explaining the processing of the extraction unit. [Figure 10] This is a diagram for explaining the processing of the precise measurement unit. [Figure 11] This is a diagram for explaining the processing of the feature point position calculation unit.

Embodiments for Carrying Out the Invention

[0014] Next, embodiments of the present invention will be described with reference to the drawings.

[0015] (Configuration) <Measurement Target Material> The measurement target material to be measured is, for example, a thick metal plate made of steel, aluminum, titanium, etc. However, the material of the measurement target material may be resin or the like. Also, for non-contact measurement, the measurement target material to be measured is not particularly limited as long as it can maintain a certain shape.

[0016] In the following embodiments, the case where the shape of the material to be measured is a thick plate having a rectangular shape (quadrilateral shape) in a plan view will be described as an example. In this specification, the plan view refers to the appearance seen from the measurement direction. In this example, the plan view will be described as the top view.

[0017] Note that the shape of the material to be measured in the plan view of the present invention is not limited to a rectangular shape, and may be other polygonal shapes, circular shapes, elliptical shapes, etc. If the reference shape (outline shape) of the material to be measured in the plan view is known in advance, characteristic points for specifying the shape can be set, so there is no problem. Further, the overall shape of the material to be measured in the present invention does not have to be flat.

[0018] That is, if the basic shape of the outer peripheral portion of the material to be measured in the plan view is known in advance, a plurality of characteristic points for specifying the shape of the material to be measured can be set, and by obtaining the positions of the plurality of characteristic points, the shape of the material to be measured can be measured. Examples of shape specification include the dimensions and profile shape of the material to be measured. In the present embodiment, a case where at least one of the shapes before and after cutting of the material to be measured made of a thick plate cut by a cutting device is the object of shape measurement will be exemplified.

[0019] <Regarding the position of the characteristic points> As shown in FIG. 1, if the measurement material 1 has a rectangular shape, at least four corner positions may be set as the positions P of the characteristic points. Further, the positions P of the characteristic points may be separately set at intermediate positions on the sides. If the measurement material 1 has a circular shape, any three or more points on the contour of the measurement material 1 are set as the positions P of the characteristic points. The positions P of the characteristic points may be set at locations where the center position and radius of the circle can be obtained. Thus, the position P of the characteristic point is set according to the reference shape, and is not limited to the above examples as long as the shape to be measured of the measurement material 1 can be specified. The location of the position P of the characteristic point may be set in advance. Note that the position P of the characteristic point is set at two or more locations. If only the width dimension of the rectangular shape is to be measured, a configuration in which the positions P of the characteristic points are set at two locations in the width direction may also be used.

[0020] In this embodiment, as shown in Figure 1, the positions of the four corners are set as the position P of the feature points. Once the positions of the four corners are identified, the dimensions and profile of the material to be measured 1 can be determined.

[0021] Furthermore, as shown in Figure 1, the region where the position P of each feature point is estimated to exist is set as the feature point existence region SA. The feature point existence region SA does not need to be rectangular in shape; it may be circular or other shapes. In this embodiment, since it is assumed that the sensor measuring within the feature point existence region SA is moved in one direction, the feature point existence region SA is set to a rectangular shape aligned with the direction of sensor movement.

[0022] The size of the feature point region SA should be set to a range that includes the range of the error based on the resolution used to extract the feature point location P, with the estimated feature point location P as the center. The feature point region SA is also called the precision measurement region.

[0023] (Shape measurement method) The shape measurement method is a method for measuring the shape of the material to be measured 1. As shown in Figure 2, the shape measurement method of this embodiment comprises a region identification step 30, a precision measurement step 31, a feature point position calculation step 32, and a dimension calculation step 33.

[0024] <Area identification process 30> The region identification step 30 performs a process to identify feature point existence regions SA, which are regions where the location P of each feature point that defines the shape of the material to be measured 1 is estimated to exist. The region identification step 30 of this embodiment comprises a rough measurement step 30A and an extraction step 30B.

[0025] [Rough measurement process 30A] The rough measurement step 30A is a process that acquires the surface shape of the entire material to be measured 1 as rough measurement data consisting of a set of three-dimensional point clouds. The surface shape is the shape as viewed from the measurement direction. The measurement of a 3D point cloud dataset can be performed using known sensors capable of measuring 3D point clouds, such as laser scanners or cameras (for image analysis).

[0026] [Extraction process 30B] The extraction step 30B performs a process to extract each feature point location region SA from the rough measurement data obtained in the rough measurement step 30A.

[0027] For example, a 3D point cloud of the height of the surface position of the material being measured 1 is extracted from the rough measurement data. However, the above height is set as a height range (threshold) with an upper limit and a lower limit, based on the height position of the material being measured 1, taking into account the error corresponding to the resolution of the sensor. Furthermore, if the surface position (top surface) of the material being measured 1 has irregularities in the height direction, this is also taken into consideration when setting the above height range.

[0028] Next, the surface shape of the material to be measured 1 is determined from the extracted 3D point cloud. In this example, the top surface of the material to be measured 1 is determined as the surface shape. Next, the corner positions of the determined surface shape are determined as the positions P of each feature point (see Figure 1). Then, the feature point existence region SA is set so as to include a range that is separated from the determined feature point position P by the measurement error of the sensor used for measurement. The measurement error is a value obtained, for example, from the distance between points in a 3D point cloud. It is preferable to set the measurement error with a wide safety margin.

[0029] As shown in Figure 3, the position P' of the feature point obtained in extraction step 30B is identified from information with low accuracy, and therefore often deviates from the position P of the true feature point. For this reason, considering the measurement error of the sensor, the feature point location SA is set based on the position P' of the feature point obtained in extraction step 30B. This ensures that the true feature point location P is reliably located within the set feature point location SA. Furthermore, although the feature point location SA is obtained from coarse information, it can be extracted as a narrow area compared to the entire upper surface of the material 1 being measured.

[0030] [others] The region identification step 30 is a process for extracting feature point existence regions SA, which are regions where the position P of each feature point is estimated to exist. In other words, it is sufficient to obtain the region where the position P of the feature point exists. Therefore, the region identification step 30 may determine the feature point existence regions SA based on information from other devices. For example, if the material to be measured 1 is a material that has been cut into a desired shape by a cutting device, the region identification step 30 may be carried out as follows.

[0031] In other words, the region identification step 30 may also set the feature point existence region SA by estimating the position of the corner based on the information of the cut position at the time of cutting. For example, the information of the position P' of the feature point is estimated from the information at the time of cutting. Then, the region that includes the range of the estimated feature point position P' plus a safety margin for the error of the information at the time of cutting is set as the feature point existence region SA.

[0032] The information at the time of cutting may be, for example, cutting position information sent to the cutting device, or cutting position information acquired during cutting. Then, the contour position of the thick plate after cutting may be estimated from the cutting information, and the position information of the corners may be obtained as feature point position information from this estimated information.

[0033] Furthermore, when the material to be measured 1 before cutting is the target, the characteristic point region SA may be identified from the transport position and transport error of a transport device such as a crane that transports the material to be measured 1 to the cutting device. However, determining the feature point region SA using the rough measurement step 30A and extraction step 30B described above allows for setting a smaller feature point region SA.

[0034] <Precision measurement process 31> The precision measurement process 31 performs a process to acquire the surface shape of the feature point-existing region SA, determined in the region identification process 30, as precision measurement data, which is a set of three-dimensional point clouds. This process is performed for each feature point-existing region SA.

[0035] The measurement of the set of 3D point clouds within the feature point region SA can be performed using known sensors capable of measuring 3D point clouds, such as laser scanners or cameras (image analysis). However, a sensor with higher resolution and a narrower measurement range than the 3D point cloud sensor used in the rough measurement process 30A should be used.

[0036] <Feature point location calculation process 32> The feature point position calculation step 32 performs a process to obtain information about the position P of feature points located in the feature point region SA from the acquired precision measurement data (see Figure 3). The calculation of the feature point position P is performed for each feature point region SA.

[0037] For example, a 3D point cloud of the height of the surface position of the material 1 being measured is extracted from the precise measurement data. However, the above height is set as a height range (threshold) with an upper limit and a lower limit set for the height of the material 1 being measured, taking into account the error according to the resolution of the sensor. Furthermore, if the surface position (top surface) of the material 1 being measured has irregularities in the height direction, this is also taken into consideration when setting the above height range. However, since the feature point existence region SA is narrow and is a corner region, irregularities can be ignored.

[0038] Next, the surface shape of the material 1 under test in the feature point region SA is determined from the extracted 3D point cloud. Then, the corner positions of the determined surface shape are determined as the positions P of each feature point. By narrowing the measurement range and performing sensing with high resolution, the position P of feature points can be determined with high accuracy.

[0039] <Dimension calculation process 33> The dimension calculation step 33 calculates the dimensions of the material to be measured 1 from the position information P of two or more feature points obtained in the feature point position calculation step 32. In this embodiment, for example, the width and length in the longitudinal direction of the thick plate (material to be measured 1) to be measured are determined as the dimensions of the material to be measured 1. The dimension calculation step 33 may also obtain profile information, which is information about the outer contour shape of the material to be measured 1.

[0040] <Regarding the rough measurement process 30A and the precision measurement process 31> Each measurement in the rough measurement process 30A and the precision measurement process 31 is performed using a measurement sensor. For example, in the rough measurement step 30A, rough measurement data is acquired using a rough measurement sensor (an example of a first measurement sensor) capable of measuring the overall surface shape of the material to be measured 1. Also, for example, in the precision measurement step 31, precision measurement data is acquired using a precision measurement sensor (an example of a second measurement sensor) which has a narrower measurement range and higher resolution than the rough measurement sensor.

[0041] (shape measuring device) The shape measuring device is a device that implements the shape measuring method described above. In this embodiment, we will explain using as an example the case in which, when a thick plate is transported to a cutting facility and cut by a cutting device to manufacture a product, the shape of the thick plate is measured by a shape measuring device at least one of the following points: before cutting and after cutting.

[0042] The purpose of measuring the shape of the thick plate before cutting is to determine the placement position of the thick plate to be cut by acquiring the position P of characteristic points on the placed thick plate. Then, based on the acquired placement position of the thick plate, the thick plate is cut into the target shape using a cutting device.

[0043] Furthermore, the purpose of measuring the shape of the cut plate is to confirm whether or not it has been cut to the target dimensions by obtaining the dimensions of the cut plate.

[0044] The shape measuring device of this embodiment comprises a coarse measuring sensor, a precision measuring sensor, and a shape calculation unit. In this embodiment, the shape calculation unit is provided in the control device. In this embodiment, we illustrate a case where a rough measurement sensor and a precision measurement sensor are provided in the movable part of the cutting device.

[0045] <Cutting device> The cutting device of this embodiment is, for example, a gantry-type cutting device, which cuts a thick plate to a target shape by, for example, gas cutting.

[0046] As shown in Figure 4, the cutting device 2 of this embodiment comprises two rails 3, a gate-type trolley 21, a width-direction drive unit 22 provided on the trolley 21, and a control device 23. A cutting device (not shown) is provided on the width-direction drive unit 22, enabling the cutting of a thick plate placed in area ARA within the two rails 3 into the target shape.

[0047] As shown in Figure 5, a surface plate 4 is provided in area ARA between the rails 3, and the material to be measured 1, made of a thick plate, is placed on the surface plate 4 in a horizontal position. In Figure 4, the surface plate 4 is omitted for clarity.

[0048] In this embodiment, when the material to be measured 1 is cut by the cutting device 2 to manufacture a product, the placement position and dimensions of the material to be measured 1 are automatically measured by the shape measuring device at least one of the following points: before cutting and after cutting.

[0049] As shown in Figure 4, the shape measuring device includes a rough measuring sensor 5, a precision measuring sensor 6, and a shape calculation unit in the control device 23, all mounted on the trolley 21 (gate-type frame) of the gate-type cutting device 2. In this embodiment, the precision measuring sensor 6 is positioned lower than the rough measuring sensor 5, that is, closer to the material to be measured 1.

[0050] As shown in Figure 4, the trolley 21 is capable of traveling on a pair of rails 3 laid in the x-axis direction (horizontal direction: approximately the longitudinal direction of the material to be measured 1) by wheels (not shown). The material to be measured 1 is placed on a surface plate 4 (omitted in Figure 4) provided between the rails 3. The trolley 21 is also provided with a width-direction drive unit 22 that can move linearly in the y-axis direction (horizontal direction perpendicular to the x-axis direction: approximately the width direction of the material to be measured 1). The width-direction drive unit 22 can be driven by a known linear motion device such as a ball screw device.

[0051] <Coarse measurement sensor 5 and precision measurement sensor 6> The coarse measurement sensor 5 and the precision measurement sensor 6 are attached to the widthwise drive unit 22 of the trolley 21. The coarse measurement sensor 5 is less expensive (lower resolution) and has a wider measurement range than the precision measurement sensor 6, and is capable of measuring the three-dimensional surface shape of the entire width of the material 1 to be measured. In this embodiment, the coarse measurement sensor 5 can be exemplified by a TOF type LiDAR sensor. The coarse measurement sensor 5 measures the distance to the material 1 by irradiating the material 1 with laser light extending along the entire width (y-axis direction) of the material 1 to be measured and measuring the time it takes for the reflected light to return. By moving this coarse measurement sensor 5 along the rail 3 in the x-axis direction using the trolley 21, the three-dimensional surface shape of the entire material 1 to be measured is measured.

[0052] The precision measurement sensor 6 has a narrower measurement range than the coarse measurement sensor 5, but has higher resolution and is a sensor capable of high-precision measurement. In this embodiment, the precision measurement sensor 6 can be exemplified as a non-contact distance sensor using the light section method. The precision measurement sensor 6 partially measures the three-dimensional surface shape of the material to be measured 1 by irradiating the material to be measured 1 with a sheet-shaped laser beam and capturing the position of the laser beam from a different angle using a camera.

[0053] In this embodiment, we will describe a case in which the surface shape of the material to be measured 1 is roughly measured by a rough measurement sensor 5 and then precisely measured by a precision measurement sensor 6. However, the same type of sensor may be used as the rough measurement sensor 5 and the precision measurement sensor 6, or a single sensor may be used to perform both rough and precision measurements. Specifically, for example, a stereo camera may be used as the rough measurement sensor 5 and the precision measurement sensor 6. Then, for example, the field of view of the stereo camera may be widened (resolution reduced) to roughly measure the entire material to be measured 1, and then the field of view of the stereo camera may be narrowed (resolution increased) to precisely measure a portion of the material to be measured 1.

[0054] <Control device 23> In this embodiment, the control device 23 is provided on the trolley 21. As shown in Figure 6, the control device 23 comprises a cutting control unit 23A and a shape calculation unit 24.

[0055] The cutting control unit 23A performs a process to cut the material to be measured 1 into the target shape while driving the trolley 21 and the width direction drive unit 22 based on the acquired dimensional shape and cutting shape information.

[0056] As shown in Figure 6, the shape calculation unit 24 includes a region identification unit 24A, a precision measurement unit 24B, a feature point position calculation unit 24C, and a dimension calculation unit 24D. In this embodiment, the region identification unit 24A includes a rough measurement unit 24Aa and an extraction unit 24Ab.

[0057] The measurement data acquired by the rough measurement sensor 5 and the precision measurement sensor 6 is then supplied to the shape calculation unit 24 of the control device 23. The extraction unit 24Ab extracts the characteristic point region SA of the material to be measured 1 from the rough measurement data acquired by the rough measurement sensor 5. The dimension calculation unit 24D calculates the dimensions of the material to be measured 1 from the precision measurement data acquired by the precision measurement sensor 6.

[0058] The region identification unit 24A executes the process of the region identification step 30. The precision measurement unit 24B executes the process of the precision measurement step 31. The feature point position calculation unit 24C executes the process of the feature point position calculation step 32. The dimension calculation unit 24D executes the process of the dimension measurement step.

[0059] In other words, the region identification unit 24A performs a process to acquire two or more feature point existence regions SA, which are regions where it is estimated that the position P of a feature point that identifies the shape of the material to be measured 1 exists, based on the measurement information of the rough measurement sensor 5.

[0060] The precision measurement unit 24B, based on the measurement information from the precision measurement sensor 6, processes the surface shape of each feature point-existing region SA, determined by the region identification unit 24A, to be acquired as precision measurement data, which is a set of three-dimensional point clouds. In other words, for each feature point-existing region SA, the precision measurement unit 24B performs sensing with the precision measurement sensor 6 to obtain precision measurement data consisting of a set of three-dimensional point clouds as three-dimensional positional information of the surface shape in each feature point-existing region SA.

[0061] The feature point position calculation unit 24C obtains information on the position P of the feature point present in the corresponding feature point present region SA from the acquired precision measurement data for each feature point present region SA.

[0062] The dimension calculation unit 24D calculates the shape of the material to be measured 1 from the position information P of two or more feature points obtained. In this embodiment, the dimension calculation unit 24D obtains the width and length of the material to be measured 1, or the profile (outer contour shape) of the material to be measured 1, from the positions P of four feature points. In other words, it is possible to obtain information about the shape of the material to be measured 1 itself and information about the position of the material to be measured 1.

[0063] <Rough measurement section 24Aa> The rough measurement unit 24Aa acquires the surface shape of the material to be measured 1 as rough measurement data, which is a collection of three-dimensional point clouds, based on the measurement values ​​from the rough measurement sensor 5.

[0064] As shown in Figure 7, the rough measurement unit 24Aa of this embodiment irradiates a laser beam from a rough measurement sensor 5 attached to the width direction drive unit 22 into the rough measurement range D1 and measures the distance to the object including the material to be measured 1. The rough measurement unit 24Aa sets the rough measurement range D1 to the entire range between the rails 3 and measures the shape of the material to be measured 1 placed between the rails 3. The rough measurement sensor 5 acquires the vertical distance to the object including the material to be measured 1 at a predetermined sampling period while the trolley 21 moves in the x-axis direction (extension direction of the rails 3) from the rough measurement start position 3A to the rough measurement end position 3B.

[0065] <Extraction part 24Ab> The extraction unit 24Ab performs a process to extract the feature point region SA in the material to be measured 1 from the rough measurement data acquired by the rough measurement unit 24Aa.

[0066] In this embodiment, the extraction unit 24Ab extracts information to identify feature point locations SA according to the steps ST1 to ST5 shown in Figure 8, and extracts four feature point locations SA from the extracted information. Then, in step ST6, it generates a movement path for the precision measurement sensor 6 to precisely measure these four feature point locations SA.

[0067] First, in step ST1, rough measurement data is acquired. Next, in step ST2, in order to extract the planar area of ​​the upper surface of the material to be measured 1 (the area of ​​the shape of the upper surface of the material to be measured 1), thresholds are set for each axis direction (x-axis, y-axis, z-axis) of the rough measurement data, and the point cloud that exists within the threshold is trimmed.

[0068] Specifically, as shown in Figure 9(a), for example, a lower limit zmin and an upper limit zmax are set in the z-axis direction, and only the point cloud within the range of zmin to zmax is extracted. This removes measurement data from the surface plate 4, etc.

[0069] Next, in step ST3, noise is removed from the rough measurement data trimmed in step ST2. Specifically, as shown in Figure 9(b), discrete point clouds, i.e., sets of points N1 where the distance between points is greater than a predetermined threshold, are removed as noise.

[0070] Next, in step ST4, a planar region defining the flat upper surface of the material to be measured 1 is detected from the rough measurement data from which noise was removed in step ST3. Specifically, as shown in Figure 9(c), in step ST3, a set of points that match the equation (model equation) of an arbitrary plane, "ax+by+cz+d=0", is obtained from the point cloud of the rough measurement data after noise removal. This removes the point cloud N2 outside the plane region as noise.

[0071] Here, since the upper surface of the material being measured 1 is horizontal or nearly horizontal, we can set the equation (model equation) of an arbitrary plane, "ax + by + cz + d = 0", so that it is horizontal or nearly horizontal. Furthermore, although this embodiment illustrates the case where the upper surface of the material to be measured 1 is flat, if the upper surface is curved, the equation of the surface defining that curved surface can be set and the point cloud of the upper surface of the material to be measured 1 can be obtained.

[0072] Next, in step ST5, the coordinates of the corners of the planar region detected in step ST4 are detected. Specifically, as shown in Figure 9(d), the outermost points of the point cloud constituting the planar region are identified, and the coordinates of the four vertices of the planar region and the edges around those vertices are calculated as "corners". This allows us to identify the location P of the feature point obtained from the rough measurement (rough information).

[0073] In this embodiment, as shown in Figure 9(d), the feature point existence region SA is defined as a range of (L / 2) in the x-axis direction from the position P of the feature point. The width of the feature point existence region SA in the y-axis direction is set to a constant fixed value. In other words, in this embodiment, each feature point existence region SA is defined using the coordinates of the position P of the feature point as a variable.

[0074] Next, in step ST6, a movement path for the precision measurement sensor 6 is generated, which passes through the four feature point locations SA defined by the coordinates of the corners detected in step ST5. In this embodiment, as described above, as shown in Figure 9(d), the range of measurement length L in the x-axis direction centered on the coordinates of each corner is defined as the range of each feature point existence region SA. In addition, in this embodiment, the width in the y-axis direction of the feature point existence region SA is set to the sensor width D2 of the precision measurement sensor 6 (see Figure 10(b)).

[0075] In this embodiment, as shown in Figure 10(b), each feature point location SA is defined by two reference points (white circle positions Q in Figure 10(b)) located at a distance of (L / 2) in the x-axis direction from the position P' of the feature point obtained in each rough measurement. Then, as shown in Figure 10(b), a path K is generated that moves through the four feature point location SAs, passing through each of the reference points Q.

[0076] <Precision measurement section 24B> In the precision measurement unit 24B, the surface shape of each of the four feature point locations SA extracted by the extraction unit 24Ab is acquired as precision measurement data, which is a collection of three-dimensional point clouds. In this embodiment, as shown in Figure 10, a precision measuring sensor 6 attached to the widthwise drive unit 22 emits laser light within the precision measuring range D2, and the distance to the object including the corners of the material to be measured is measured by imaging the position of the laser light emission.

[0077] Specifically, the movement of the trolley 21 and the width-direction drive unit 22 is controlled as shown by the arrows in Figure 10(b) so that the center of the precision measurement range D2 of the precision measurement sensor 6 moves along the path K determined in step ST6. At this time, the precision measurement sensor 6 supplies the measured value measured while the precision measurement range D2 moves between two reference points Q to the shape calculation unit 24. In this way, the four feature point existence regions SA are measured with a measurement length L. As described above, in this embodiment, each feature point existence region SA is defined by two reference points Q separated by a measurement length L.

[0078] <Feature point position calculation unit 24C, dimension calculation unit 24D> In the feature point position calculation unit 24C and the dimension calculation unit 24D, the precise position P of the feature point is detected in steps ST10 to ST15 shown in Figure 11, and the dimensions of the material to be measured 1 are calculated from the precise position P of the feature point.

[0079] Here, steps ST10 to ST14 are the processing of the feature point position calculation unit 24C, and the processing of step ST15 corresponds to the processing of the dimension calculation unit 24D. Since the processes in steps ST10 to ST14 are the same as those in steps ST1 to ST5, a detailed explanation will be omitted.

[0080] However, instead of processing the entire surface shape of the upper surface of the material being measured 1, the processing is performed for each feature point region SA. That is, for each feature point region SA, the planar region is detected again based on the precise measurement data, and the coordinates of the corners (vertices and edges) of the planar region are precisely recalculated.

[0081] Then, in step ST15, which is the processing of the dimension calculation unit 24D, the distance between the coordinates of the corners is calculated, and from this distance, the length dimension and width dimension of the material to be measured 1 are calculated. Note that "distance between the coordinates of the corners" may refer to the distance between vertices, or the distance between edges around a vertex.

[0082] (Product manufacturing method) In this embodiment, when manufacturing a product by cutting the material to be measured 1 with the gantry-type cutting device 2, shape measurement is performed using the shape measurement method of the shape measuring device described above, at least one of the following: before cutting and after cutting.

[0083] According to the above configuration, when measuring the shape of the material to be measured 1, a rough measurement sensor 5 with a wide measurement range is used to pre-measure the placement position and approximate surface shape of the material to be measured 1. Subsequently, based on the rough measurement data, a precision measurement sensor 6 with high resolution is used to measure the dimensions and placement position of the material to be measured 1. Therefore, the dimensions of the material to be measured 1 can be measured automatically without human intervention.

[0084] Furthermore, compared to measuring the dimensions of the material to be measured 1 using only a wide-area sensor with low resolution, this embodiment can improve the accuracy of dimensional measurement. On the other hand, compared to measuring the dimensions of the material to be measured 1 using only a narrow-area sensor with high resolution, this embodiment can efficiently identify the placement position of the material to be measured 1 and measure its dimensions.

[0085] Furthermore, by measuring the dimensions of the material to be measured before and after cutting, it becomes possible to manufacture products with high dimensional accuracy.

[0086] (others) This disclosure may also take the following form: (1) A method for measuring the shape of a material to be measured, A region identification step involves acquiring two or more feature point existence regions, which are regions where it is estimated that feature points exist that identify the shape of the material to be measured, and A precision measurement process is performed to acquire the surface shape of each identified feature point region as precision measurement data consisting of a set of 3D point clouds, A feature point position calculation process is performed to obtain the positional information of the above feature points located in each feature point region from the acquired precision measurement data. A method for measuring the shape of a material to be measured, comprising the following components. (2) A dimension calculation step is included in which the dimensions of the material to be measured are calculated from the positional information of two or more characteristic points obtained. (3) The above area identification process is: A rough measurement step is performed to acquire the surface shape of the material to be measured as rough measurement data consisting of a collection of three-dimensional point clouds. From the rough measurement data obtained in the above rough measurement step, an extraction step is performed to extract regions where two or more feature points exist. It is equipped with. (4) In the rough measurement step described above, the rough measurement data is acquired using a first measuring sensor capable of measuring the overall surface shape of the material to be measured. In the precision measurement process described above, the precision measurement data is acquired using a second measurement sensor that has a narrower measurement range and higher resolution than the first measurement sensor. (5) The material to be measured is a plate, In the extraction process described above, a planar region of the upper surface of the material to be measured is extracted from the rough measurement data, the positions of two or more feature points are estimated from the extracted planar region, and the region containing the position of each estimated feature point is extracted as the feature point presence region. (6) The material to be measured is a polygonal plate, The above-mentioned feature points are defined as the positions of the corners of the polygonal shape. (7) The material to be measured is a material that has been cut into the desired shape using a cutting device. In the above-mentioned region identification process, the region where feature points exist is identified based on the information of the cutting position during the above-mentioned cutting. (8) A shape measuring device for measuring the shape of a material to be measured, A region identification unit that acquires two or more feature point existence regions, which are regions where it is estimated that feature points exist that identify the shape of the material to be measured, A precision measurement unit acquires the surface shape of each acquired feature point region as precision measurement data consisting of a set of 3D point clouds, A feature point position calculation unit obtains the positional information of the feature points located in the corresponding feature point region from the acquired precision measurement data, It is equipped with. (9) The device includes a dimension calculation unit that calculates the dimensions of the material to be measured from the positional information of two or more feature points obtained. (10) The above-mentioned region identification unit is A rough measurement unit acquires the surface shape of the material to be measured as rough measurement data consisting of a collection of three-dimensional point clouds, An extraction unit extracts regions containing two or more feature points from the rough measurement data acquired by the rough measurement unit described above, It is equipped with. (11) The rough measurement unit acquires the rough measurement data using a first measurement sensor capable of measuring the overall surface shape of the material to be measured. The precision measurement unit acquires the precision measurement data using a second measurement sensor that has a narrower measurement range and higher resolution than the first measurement sensor. (12) The material to be measured is a material that has been cut into the desired shape using a cutting device. The first and second measurement sensors are installed in the cutting device. (13) A method for manufacturing a product by processing the above-mentioned material to be measured, The shape of at least one of the materials to be measured, both before and after processing, is measured using the shape measuring device of the present disclosure. Product manufacturing method. [Examples]

[0087] The surface shape of the thick plate was measured using the shape measuring device described in the embodiment. For the thick plates, we used rectangular steel plates measuring 1215mm in width and 4375mm in length. Furthermore, the coarse measurement sensor 5 uses a LIDAR sensor that measures the shape of an object using the TOF method, while the precision measurement sensor 6 uses a laser profiler that measures the shape of an object using the light section method.

[0088] As a result, we confirmed that the generation of the four feature point regions SA and the movement path K of the precision measurement sensor was successful. Furthermore, when comparing the precise measurements of the thick plate using the device of this embodiment with the dimensions measured manually with a measuring tape in the longitudinal and width directions, the maximum error with the manual measurement was 1.8 mm, indicating that the dimensions can be calculated with good accuracy. [Explanation of symbols]

[0089] 1 Material to be measured 2 Cutting device 3 rails 5. Coarse measurement sensor 6. Precision measuring sensors 21 bogies 22 Width direction drive unit 23 Control device 23A Cutting Control Unit 24 Shape calculation section 24A Area identification part 24Aa Rough measurement section 24Ab extraction part 24B Precision measurement section 24C Feature Point Position Calculation Unit 24D Dimension Calculation Unit 30 Area identification process 30A rough measurement process 30B Extraction process 31 Precision measurement process 32 Feature Point Position Calculation Process 33 Dimension Calculation Process D1 Rough measurement range D2 Precision Measurement Range K path Q reference point SA Feature Point Existence Region

Claims

1. A shape measurement method for measuring the shape of a material to be measured, A region identification step involves acquiring two or more feature point existence regions, which are regions where it is estimated that feature points exist that identify the shape of the material to be measured. A precision measurement process is performed to acquire the surface shape of each identified feature point region as precision measurement data consisting of a set of three-dimensional point clouds, A feature point position calculation process is performed to obtain the positional information of the above feature points located in each feature point region from the acquired precision measurement data. A method for measuring the shape of a material to be measured, comprising the following components.

2. A dimension calculation process that calculates the dimensions of the material to be measured from the positional information of two or more characteristic points obtained. A method for measuring the shape of a material to be measured according to claim 1, comprising:

3. The above region identification process is, A rough measurement step is performed to acquire the surface shape of the material to be measured as rough measurement data consisting of a collection of three-dimensional point clouds. From the rough measurement data obtained in the above rough measurement step, an extraction step is performed to extract regions where two or more feature points exist. A method for measuring the shape of a material to be measured, as described in claim 1, comprising:

4. In the above rough measurement step, the rough measurement data is acquired using a first measurement sensor capable of measuring the overall surface shape of the material to be measured. In the precision measurement process described above, the precision measurement data is acquired using a second measurement sensor that has a narrower measurement range and higher resolution than the first measurement sensor. A method for measuring the shape of a material to be measured, as described in claim 3.

5. The material being measured is a plate, In the extraction process described above, a planar region of the upper surface of the material to be measured is extracted from the rough measurement data, the positions of two or more feature points are estimated from the extracted planar region, and the regions containing the positions of each estimated feature point are extracted as the feature point presence regions. A method for measuring the shape of a material to be measured, as described in claim 3.

6. The material being measured is a polygonal plate. The above characteristic points are defined as the positions of the corners of the polygonal shape. A method for measuring the shape of a material to be measured, as described in claim 1.

7. The material to be measured above is a material that has been cut into the desired shape using a cutting device. In the above region identification step, the region where feature points exist is identified based on the information of the cutting position during the above cutting. A method for measuring the shape of a material to be measured, as described in claim 1.

8. A shape measuring device for measuring the shape of a material to be measured, A region identification unit that acquires two or more feature point existence regions, which are regions where it is estimated that feature points exist that identify the shape of the material to be measured, A precision measurement unit acquires the surface shape of each acquired feature point region as precision measurement data consisting of a set of three-dimensional point clouds, A feature point position calculation unit obtains the positional information of the feature points located in the corresponding feature point region from the acquired precision measurement data, A shape measuring device for a material to be measured, equipped with the following features.

9. The shape measuring device for a material to be measured according to claim 8, further comprising a dimension calculation unit that calculates the dimensions of the material to be measured from the positional information of two or more characteristic points obtained.

10. The above-mentioned region identification unit is, A rough measurement unit acquires the surface shape of the material to be measured as rough measurement data consisting of a collection of three-dimensional point clouds, An extraction unit extracts regions containing two or more feature points from the rough measurement data acquired by the rough measurement unit described above, A shape measuring device for a material to be measured according to claim 8, comprising:

11. The rough measurement unit acquires the rough measurement data using a first measurement sensor capable of measuring the overall surface shape of the material to be measured. The precision measurement unit acquires the precision measurement data using a second measurement sensor that has a narrower measurement range and higher resolution than the first measurement sensor. A shape measuring device for a material to be measured, as described in claim 10.

12. The material to be measured above is a material that has been cut into the desired shape using a cutting device. The first and second measuring sensors are installed in the cutting device. A shape measuring device for a material to be measured, as described in claim 11.

13. A method for manufacturing a product by processing the above-mentioned material to be measured, The shape of at least one of the materials to be measured, both before and after processing, is measured using the shape measuring device described in any one of claims 8 to 12. Product manufacturing method.

Images