JP6360706B2 - Shape measuring apparatus, program, recording medium, and method - Google Patents

Description

  The present invention relates to a shape measuring device, a program, a recording medium, and a method used for measuring the shape of a coil spring.

  Conventionally, Patent Document 1 describes a coil spring shape measuring device and a shape measuring method. In this apparatus and method, a laser that irradiates the surface of a coil spring with slit light spreading in the axial direction of the coil spring, a rotary stage that rotates the coil spring around its axis, and the reflected light from the surface of the coil spring are photographed. A CMOS camera and a computer that calculates the coordinate value of the coil spring wire from the captured images taken by the CMOS camera for each of a plurality of rotation angles when the coil spring is rotated about its axis.

  According to this shape measuring apparatus, the position of the coil spring can be specified from the captured image of the coil spring, and the shape of the coil spring can be measured.

  However, in this apparatus and method, since the coordinate value of the coil spring wire is calculated from the photographed image taken by the CMOS camera, image processing, erasure of the reflected image of the fixing jig, and the like are required, and the processing becomes complicated. There was a problem.

JP 2010-101893 A

  The problem to be solved is that the shape of a coil spring or the like can be measured, but the processing is complicated by image processing or the like.

The shape measuring device of the present invention is a shape measuring device having a function of measuring the shape of an object having an arc in a cross-sectional outer shape in order to facilitate processing without requiring image processing, Laser displacement sensor for detecting coordinate values of measurement points that irradiate and reflect laser light from one side of the cross-sectional outer shape at a plurality of measurement points along one side of the cross-sectional outer shape, and coordinates of the plurality of measurement points A calculation unit that approximates the shape of the arc from a value group, and the calculation unit obtains a differential value from each coordinate value and compares them with each other to determine the maximum and minimum differential values of the coordinate value group corresponding to the arc. Both end points of a coordinate value group corresponding to the arc are obtained based on the value determination, and the approximation is performed from coordinate values continuous between the end points.

The shape measurement program of the present invention is a shape measurement program that causes a computer to realize a function of measuring the shape of an object having an arc in a cross-sectional outer shape, and the function is configured to emit laser light from one side of the cross-sectional outer shape. A calculation function for approximating the shape of the arc from the coordinate value group detected at a plurality of measurement points along one side of the cross-sectional outer shape, the coordinate value of the measurement point that irradiates and reflects, the calculation function, The differential value in the coordinate value is obtained and compared with each other, and the end points of the coordinate value group corresponding to the arc are obtained based on the maximum / minimum value determination of the differential value of the coordinate value group corresponding to the arc. The approximation is performed from coordinate values that are continuous with each other.

The shape measurement program recording medium of the present invention is a computer-readable shape measurement program recording medium recording a shape measurement program for causing a computer to realize a function of measuring the shape of an object having an arc in a cross-sectional outer shape. The function is to calculate the shape of the arc from a group of coordinate values detected at a plurality of measurement points along one side of the cross-sectional outer shape by radiating laser light from one side of the cross-sectional outer shape and reflecting it. An approximation calculation function is provided, and the calculation function corresponds to the arc based on the determination of the maximum value / minimum value of the differential value of the coordinate value group corresponding to the arc by obtaining a differential value at each coordinate value and comparing each other. Both end points of the coordinate value group to be obtained are obtained, and the approximation is performed from coordinate values continuous between the both end points.

The shape measurement method of the present invention is a shape measurement method for measuring the shape of an object having an arc in a cross-sectional outer shape, and is a measurement point at which laser light is irradiated and reflected from one side of the cross-sectional outer shape by a laser displacement sensor And a calculation step of approximating the shape of the arc from the coordinate value group of the plurality of measurement points. The step is to obtain a differential value at each coordinate value and compare with each other to obtain both end points of the coordinate value group corresponding to the arc based on the maximum value / minimum value determination of the differential value of the coordinate value group corresponding to the arc, The approximation is performed from the coordinate values continuous between the two end points.

  Since the shape measuring apparatus of the present invention has the above-described configuration, the coordinate value of the measurement point that is irradiated and reflected from one side of the cross-sectional outer shape is detected at a plurality of measurement points along one side of the cross-sectional outer shape. The shape of the arc can be approximated from the coordinate value group of the plurality of measurement points.

  In this case, the differential value at each coordinate value is obtained and compared with each other to obtain the end points of the coordinate value group corresponding to the arc, and the arc shape can be approximated from the coordinate values continuous between the end points.

  For this reason, image processing is not required, and processing can be performed easily and reliably. In addition, since the differential value of each coordinate value is used, it is not necessary to know both end points of the arc as a processing reference, the processing is simple, and a linear coordinate value group such as a jig is detected. Can be processed as is.

  Since the shape measurement program of the present invention has the above configuration, the above functions can be realized by a computer, and the processing can be performed easily and reliably without performing image processing. In addition, since the differential value of each coordinate value is used, it is not necessary to know both end points of the arc as a processing reference, the processing is simple, and a linear coordinate value group such as a jig is detected. Can be processed as is.

  Since the shape measurement program recording medium of the present invention has the above-described configuration, the above functions can be read and realized by a computer, and the processing can be easily and reliably performed without performing image processing. In addition, since the differential value of each coordinate value is used, it is not necessary to know both end points of the arc as a processing reference, the processing is simple, and a linear coordinate value group such as a jig is detected. Can be processed as is.

  Since the shape measuring method of the present invention has the above-described configuration, the processing can be performed easily and reliably without image processing. In addition, since the differential value of each coordinate value is used, it is not necessary to know both end points of the arc as a processing reference, the processing is simple, and a linear coordinate value group such as a jig is detected. Can be processed as is.

It is a composition schematic diagram of a shape measuring device. Example 1 It is a schematic diagram of image data. Example 1 It is a schematic diagram of the detection data of a laser displacement sensor. Example 1 It is a front view which shows an example of a coil spring. Example 1 It is explanatory drawing which shows the concept of the isolation | separation process of an adjacent circular arc. Example 1 It is explanatory drawing which shows the data of a coordinate value group. Example 1 It is explanatory drawing which shows circular regression. Example 1 It is explanatory drawing which shows the example of a measurement of the close_contact | adherence winding part of a coil spring. Example 1 It is explanatory drawing which shows the extraction concept of the starting point / end point candidate of an arc. Example 1 It is explanatory drawing which shows the extraction concept of a circular arc. Example 1 It is a flowchart of the calculation process in shape measurement. Example 1 This is a subroutine for starting point candidate extraction. Example 1 This is an end point candidate extraction subroutine. Example 1 This is an arc candidate extraction subroutine. Example 1 It is a composition schematic diagram of a shape measuring device. (Example 2)

  As shown in FIG. 1, a shape measuring device 1 having a function of measuring the shape of a coil spring 3 having a circular arc in a cross-sectional outer shape is used for the purpose of making it possible to simplify processing without requiring image processing. The laser displacement sensor 7 that detects the coordinate values of the measurement points that are irradiated and reflected from one side of the cross-sectional outer shape at a plurality of measurement points along one side of the cross-sectional outer shape, and the coordinates of the plurality of measurement points A computing unit 13 for approximating the shape of the arc from the value group. The computing unit 13 obtains a differential value at each coordinate value and compares them with each other to obtain end points of the coordinate value group corresponding to the arc. This was realized by performing the approximation from coordinate values consecutive in between.

[Shape measuring device]
FIG. 1 is a schematic configuration diagram of a shape measuring apparatus.

  As shown in FIG. 1, the shape measuring apparatus 1 according to the first embodiment of the present invention approximates an arc of a wire forming the coil spring 3 as an object having an arc in a cross-sectional outer shape, and detects the center position thereof. The coil shape of the coil spring 3 is measured.

  The shape measuring apparatus 1 includes a jig 5 that supports and fixes a coil spring 3 that performs detection, a movable unit 9 that supports a laser displacement sensor 7, a control unit 11, and a calculation unit 13. The calculation result of the calculation unit 13 is output to the display unit 15, and the coil shape and the like of the calculation result are displayed.

  In the jig 5, an upper cone 5c is fitted on an upper portion of a central core 5b fixed to the lower cone 5a, and the coil spring 3 is supported in a fixed posture by upper and lower cones 5c, 5a. This jig 5 is fixed on a stage 17 that does not rotate.

  The movable part 9 includes a rotary table 23 and a vertical drive mechanism 25.

  The rotary table 23 is arranged horizontally and is coupled to a table rotary motor 27. The table rotation motor 27 is supported and fixed to the fixed table 21.

  The vertical drive mechanism 25 is attached to the outer peripheral portion of the rotary table 23. The vertical movement mechanism 19 of the vertical drive mechanism 25 is supported so that the laser optical axis of the laser displacement sensor 7 faces the rotation axis of the rotary table 23 and is driven to move up and down.

  Further, the laser displacement sensor 7 pivots around the jig 5 (coil spring 3) together with the vertical drive mechanism 25 by the rotational drive of the rotary table 23.

  The laser displacement sensor 7 performs measurement at a preset time interval. The measured value is synchronized with the position information of the rotary table 23 and the vertical drive mechanism 25 by the control unit 11 and is output to the calculation unit 13 as the coordinate value of the measurement point.

  The control unit 11 is configured by a computer and includes a CPU, a ROM, a RAM, and the like.

  By the drive, the laser displacement sensor 7 linearly moves from the lower side to the upper side in the vertical direction (the vertical direction in FIG. 1, the coil axis direction) along one side of the circular cross-sectional shape of the wire 3 a of the coil spring 3. .

  By this relative movement, the coordinate value of the measurement point that irradiates and reflects the laser beam to the outer surface of the wire 3a from one side in the coil radial direction, which is one side of the outer diameter shape of the cross section, is detected. This coordinate value is input to the calculation unit 13.

  A coordinate value group of a plurality of measurement points along one side of the cross-sectional outer shape corresponds to a substantially semicircle of the circular cross section of the wire 3a of the coil spring 3 in this embodiment. Detection by the laser displacement sensor 7 obtains a plurality of coordinate value groups in the coil axis direction corresponding to a substantially semicircle having a circular cross-sectional shape of the wire 3a by movement in the coil axis direction.

  In addition, since the laser displacement sensor 7 moves around the coil spring 3 by the rotation of the rotary table 23 by the drive of the table rotation motor 27, the coordinate value group substantially corresponding to the semicircle is 360 with respect to the coil spring 3. A plurality of groups are detected along the coil shape of the coil spring 3 in the range of ° and input to the calculation unit 13.

  The calculation unit 13 is configured by a computer and includes a CPU, a ROM, a RAM, and the like. The calculation unit 13 can be configured by any of the same or different computers as the control unit 11. The calculation unit 13 is installed with a shape measurement program to be described later.

  The shape measurement program causes a computer to realize the function of measuring the shape of an object having an arc in the cross-sectional outer shape. In this embodiment, the cross-sectional outer shape of the wire 3a of the coil spring 3 is as shown in the flowchart described later. A semicircle is approximated as a circular arc, and a circle is approximated from the semicircle to obtain the center position thereof, and the coil shape is calculated.

  The shape measurement program can be realized by using a shape measurement program recording medium on which the shape measurement program is recorded and causing the calculation unit 13 configured by a computer to read the program.

  The calculating part 13 approximates the circular cross-sectional shape of the wire 3a of the coil spring 3 from the coordinate value group of a some measurement point as mentioned later. The cross-sectional center position is calculated from this cross-sectional shape. The cross-sectional center position is calculated along the entire coil shape of the wire 3a. This calculation is performed in a plurality of groups for each coordinate value group, and the coil shape along the wire 3a is calculated.

  The display unit 15 visually outputs and displays the relationship between the height of the coil spring, the number of coil turns, and the coil center radius depending on the calculation result of the calculation unit 13, or the shape of the coil spring is visually displayed as a continuous center of the cross section. Output is displayed. These results may be printed out.

  Therefore, the laser displacement sensor 7 is driven together with the rotation of the rotary table 23 and the drive of the vertical drive mechanism 25 under the control of the control unit 11, and the outer diameter of the cross section along one side of the cross section of the wire 3a of the coil spring 3 is irradiated by the laser. Data of a plurality of measurement points according to the shape is acquired.

From such data, the calculation unit 13 detects a coordinate value group corresponding to a substantially semicircle in a specific cross section of the wire 3a. A plurality of groups of coordinate value groups are detected along the coil shape of the wire 3a. In each coordinate value group, a circular cross-sectional shape can be approximated as will be described later, and the center position of the circular cross-section of the wire 3a can be continuously detected along the coil shape.
[Detection features]
The shape measuring device, program, recording medium, and method used for measuring the shape of the coil spring according to the first embodiment do not use a photographed image photographed with a CMOS camera as data as in Patent Document 1, but a laser displacement. Sensor detection data is used.

  FIG. 2 is a schematic diagram of image data, and FIG. 3 is a schematic diagram of detection data of a laser displacement sensor.

  The input data in Patent Document 1 is image data obtained by photographing reflected light as shown in FIG. In this image data, the measurement points are in square pixel units, and all the grids are filled with luminance, so the continuity and discontinuity of the points are clear and the end points of the arc are easy to detect. There's a problem.

  That is, when measuring the three-dimensional coordinates of an object using a general optical system camera, the image distance of the camera optical system, the intersection position of the imaging surface and the lens optical axis, the accurate optical system of the laser light source and the camera optical system are accurate. Accurate values such as positional relationship and distortion amount including lens distortion are required.

  Since these values are all difficult to directly measure, usually, a calibration board on which dots or the like having a constant pitch are printed is photographed at various distances and angles, and calculated from the images. If this calibration is not performed accurately, a correct measurement result cannot be obtained.

  In general, the distance and angle of shooting on the calibration board are often determined by the operator at random, and the reproducibility is poor, and it is difficult to calibrate appropriately. It is also difficult to confirm whether calibration has been performed accurately.

  Furthermore, since the whole is photographed at once, the resolution is generally inferior compared with a laser displacement sensor that performs measurement one by one.

  On the other hand, in Example 1, since the measurement point is taken with the detection data of the laser displacement sensor as shown in FIG. 3, the resolution is high, but the obtained point is discrete with respect to the resolution of the coordinates, so it is continuous or discontinuous. Although there is no clear concept, it is necessary to judge the continuity and detect the end point itself, but it has the following advantages.

  That is, the processing by the laser displacement sensor 7 is not an image processing for irradiating and shooting a laser beam on the axis, but a measurement by a displacement meter using a spot-like laser, and can be stably performed regardless of camera calibration or the like. Higher accuracy can be expected.

  In addition, since the measurement of the apparatus according to the first embodiment is spot-like, it is possible to measure not only in the axial direction of the coil spring 3 but also in the rotational direction, and to detect the spring terminal with high accuracy.

  FIG. 4 is a front view showing an example of a coil spring. The coil spring 3 has tightly wound portions 3b and 3c at both ends, and also separates and extracts the tightly wound portions 3b and 3c.

  In the first embodiment, the detection and separation of the end points are simultaneously performed using the inclination of the tangent line of the arc. Even if the coordinates are close and continuous, it can be detected as an end point.

  FIG. 5 is an explanatory diagram illustrating the concept of adjacent arc separation processing.

  Regarding the separation processing of adjacent arcs, in Patent Document 1, cos∠PAB and cos∠PBA are used depending on the angle connecting the end points A and B of the contour line and each point P as shown in FIG. On the other hand, in the first embodiment, the tangent slope at the point P is used. In addition, R shows a coil radial direction and Z shows a coil height direction.

  For this reason, in Patent Document 1, it is necessary to obtain both end points A and B of the arc in advance for the arc separation. Further, it is assumed that data to be processed does not include anything other than an arc representing a wire.

  Therefore, in Patent Document 1, it is necessary to input the shape dimension of a jig in advance in order to remove a portion indicating an irrelevant jig.

On the other hand, in the first embodiment, even if the points A and B are unknown, the differential value of the curve obtained only by the points near the point P, that is, the slope of the tangent line is used. Since the maximum value and minimum value of a series of inclinations are seen, processing can be performed as it is even if a linear shape that does not change inclination like a jig is mixed. For this reason, it is not necessary to input the shape of the jig, and even when a plurality of jigs are used, no setup change is required.
[Calculation process overview]
(Overview)
FIG. 6 is an explanatory diagram showing data of a coordinate value group. The horizontal axis in FIG. 6 is the radius r / mm from the coil center, and the vertical axis is the coil height z / mm in the vertical direction.

  The data measured as described above is a point group including the radial displacement and height of the surface of the wire at the outer periphery of the spring at a certain rotational coordinate as shown in FIG. In FIG. 6, dots are thinned out for easy understanding. The obtained data includes a lot of data such as jig parts other than the wire part, and these are not distinguished as data. Each arc portion represents the surface of the spring wire, the shaded portion represents the upper and lower cones 5c and 5a on which the spring is installed, and the vertical line portion represents the central core 5b (FIG. 1).

  Each arc represents a part of the wire cross section at a certain coordinate. By obtaining the center of this arc, the r coordinate and the z coordinate are known, and the solid coordinate of the wire cross section center is obtained. In addition, since the wire has a pitch and is not horizontal, it is scanned obliquely and is strictly an elliptical arc. Of course, the wire cross-section center coordinates may be obtained using elliptical regression, but circular regression is a sufficient approximation with standard coil spring pitch, and there are fewer regression variables compared to elliptical approximation. The results are more stable and more stable regression results can be obtained. For this reason, circular approximation is used in Example 1, but elliptical approximation may be used in the measurement of a coil spring having a large pitch.

  The center coordinates of the wire cross section can be obtained by performing circular regression after removing unnecessary data and separating the point cloud into those showing only the respective arc portions. This operation is repeated at each θ coordinate in the coil circumferential direction, and then the wire cross-sectional centers obtained at each θ coordinate are arranged in turn order, so that the three-dimensional coordinates of the wire cross-sectional center at each number of turns of the spring can be obtained, and the three-dimensional shape of the spring Is obtained.

(Preprocessing)
The data measured as described above is
• Sampling is not necessarily equally spaced with respect to the vertical coordinate.

    -Sampling may be performed on the same vertical coordinate, and duplication may occur.

・ Especially due to servo overshoot and hunting at the start and end points, it does not always increase monotonously with respect to the vertical coordinates.
It has the characteristics such as.

  For this reason, since it is difficult to perform the processing as it is, preprocessing is performed so that the obtained data satisfies the following.

    ・ No vertical coordinate overlap.

  -Monotonically increasing with respect to vertical coordinates.

  -Does not include abnormal points in the wire position.

  The pre-processing procedure is as follows.

  1. Remove outliers.

  2. Sort data for vertical coordinates.

  3. If there are multiple displacement data at the same position, the average value, the mode value, or the representative value such as the first and last in terms of time is taken as the displacement at that position.

(Extraction of wire part)
Pay attention to the slope of the graph in order to separate the continuous arcs of closely wound windings. In order to measure the displacement outside the wire, the measurement points are arranged in a semicircular arc shape in the wire. Therefore, the displacement at the wire portion first has a gradient in the maximum increasing direction, and the gradient decreases monotonously and gradually becomes gentle. It becomes 0 at the apex, then starts decreasing, and finally reaches the maximum decreasing gradient. Therefore, in a series of changes in inclination, the maximum value is the starting point of the arc and the minimum value is the end point.

  Next, the slope of the measurement point is obtained by numerical differentiation. Since the obtained data is not smooth due to the surface roughness and noise of the wire, large disturbance occurs when numerical differentiation is performed at adjacent points as it is. In order to obtain a smooth differential value, the measurement points are smoothed in advance. For smoothing, for example, moving average, convolution of functions, or the like is used. The smoothed value is used only for numerical differentiation and not for circular regression described later.

  Next, using this differential value, a candidate point for the starting point of the arc portion is obtained. The arc has the largest differential value at the start point, the value increases in the vertical axis direction of the graph up to the vertex, the differential value is positive, the differential value becomes 0 at the vertex, and the differential value is negative because it decreases until the end point. The differential value is the smallest at the end point.

  A point having a maximum differential value and having a value equal to or greater than the minimum threshold value among consecutive point groups having positive differential values equal to or less than the maximum threshold value is set as a candidate point for the start point. In the region where the differential value near the apex is small, the differential value is switched between positive and negative due to variations, so that an increase / decrease is determined so that each point group overlaps.

  Candidate points for the end point of the arc are reversed by reversing the sign of the threshold value and performing the same operation. A point having a minimum differential value and not more than the maximum threshold value among consecutive points having negative differential values not less than the minimum threshold value is set as a candidate point for the end point.

A group of points from adjacent start point candidates to end point candidates among the start point candidates and end point candidates are defined as arc candidates. The difference in the vertical coordinate from the start point to the end point of the arc candidate point group is roughly the length of the chord, and when this is not within the threshold expected from the known wire diameter, this candidate is excluded.
(Circular regression)
FIG. 7 is an explanatory diagram showing circular regression. The horizontal axis in FIG. 7 is the radius r / mm from the coil center, and the vertical axis is the coil height z / mm in the vertical direction.

  The center coordinates are obtained by returning to the circle from the point group constituting the arc obtained in the previous section. In Example 1, circular regression by the least square method is used. Returning to a circle is equivalent to finding the center coordinates and radius.

  Furthermore, in Example 1, when the regression error with respect to a regression result becomes more than a threshold value, it will exclude as not being a circular arc.

  FIG. 7 shows the actual measurement data processed. The thick data is a series of points extracted as arcs (point group), and the thin circle represents a circle regressed from the point group. Points were correctly extracted and processed.

(Number of coil turns)
Each obtained circle corresponds to a cross section of the wire. Find which number of turns on each winding spring corresponds to each circle. For an arc detected at a certain rotational coordinate, the rotational coordinate is treated as a reference for the number of turns, and thereafter, the section is treated as a cross section with a large number of turns by 360 degrees in ascending order of vertical coordinates. In the left-handed spring, the winding direction and the rotation coordinates are opposite, so the coordinates are reversed.

  In this method, if an arc cannot be detected due to noise or the like on the way, the number of turns of the subsequent arc is shifted by 360 degrees. The data that could not be detected shall be no data.

  Thus, the cylindrical coordinates of the center of the wire at each number of turns are obtained, and the spring shape is obtained.

(result)
FIG. 8 is an explanatory view showing a measurement example of the tightly wound portion of the coil spring. The horizontal axis in FIG. 8 is the radius r / mm from the coil center, and the vertical axis is the coil height z / mm in the vertical direction. In FIG. 8, the black dot data is a collection of points used for the regression, and the thin circle is the regression result.

  The continuous arc of the tightly wound portion could be accurately separated by the algorithm using the inclination of Example 1. That is, a shape measuring device using a laser displacement sensor can be applied to the coil spring.

Further, the vicinity of the start point and end point of the arc has a large inclination with respect to the laser optical axis, the amount of reflected light is reduced, and the accuracy is easily lowered. In the algorithm of the first embodiment, an end portion having a large inclination can be effectively removed, and an improvement in accuracy can be expected.
[Concept of extracting start / end points]
FIG. 9 is an explanatory diagram showing a concept of extracting a starting point / ending point candidate of an arc. In FIG. 9, the horizontal axis is the coil height z / mm in the vertical direction, the left side of the vertical axis is the radius r / mm from the coil center, and the right side of the vertical axis is the differential value indicating the slope of each point on the arc ( differential).

  In FIG. 9, the upper two arcs are the coordinate value groups AC1 and AC2 of the measurement points of the tightly wound portion, and the upper inclined line is the coordinate value group L of the measurement points at the lower cone 5a of the jig 5. is there. In the figure, the collection of points in the middle row shows the differential value of each measurement point.

  In the algorithm of the first embodiment, points where differential values are continuous within the grouping range GR are grouped as one group. The threshold of the grouping range is set from the increase / decrease determination threshold to the maximum threshold, for example, −0.5 to 4, and the points where the differential values are continuous within this threshold are recognized as a group. In FIG. 9, the differential values that are continuously within the threshold value of the grouping range GR are grouped into groups G1, G2, G3, G4, and G5.

  Within each group G1, G2, G3, G4, G5, the point having the maximum differential value is recognized as a starting point candidate (starting point candidate) IP1, IP2 which is one end point of the arcs AC1, AC2. However, if the maximum differential value of the group is less than the minimum threshold, for example, 0.8, it is not considered as a starting point candidate. Therefore, the endpoint adoption range PR is from the minimum threshold to the maximum threshold, and the start point candidates IP1 and IP2 are recognized as the maximum value in the endpoint adoption range PR.

  The extraction of the end point is the same process. However, since the sign of the differential value becomes negative, the grouping that is decreasing in the negative direction is regarded as a group, and the end point candidate is recognized as the minimum value.

The maximum threshold and the minimum threshold are ideally within the range of the maximum angle of the circular arc to be extracted, but in practice ・ There are limitations on the maximum inclination angle that can be measured by the laser displacement meter. Do not pick up a large gradient other than an arc due to a step such as a step ・ Select a threshold in consideration of not picking up a large gradient due to abnormal values, outliers, etc.

  The increase / decrease determination threshold value is ideally 0, which is the boundary between increase and decrease.

  Actually, if it is set to 0, the linear part such as the center of the arc and the straight part of the cone, the core metal, etc. will fluctuate with a value near small zero, and there will be many small groups there, so consider the actual measured value from the viewpoint of processing stability. Thus, the threshold value has some allowance. Even if it is zero, there is no problem because the value of the small group does not fall within the endpoint adoption range.

As the differential value, a numerical differential value obtained by quadratic Lagrange complementation is used in the embodiment. The arc is used because it is expressed by a quadratic expression, but any numerical differentiation method can be used.
[Arc extraction concept]
FIG. 10 is an explanatory diagram showing the concept of arc extraction. In FIG. 10, the horizontal axis is the coil height z / mm in the vertical direction, the left side of the vertical axis is the radius r / mm from the coil center, and the right side of the vertical axis is the differential value indicating the slope of each point on the arc ( differential).

  In FIG. 10, the upper two arcs are the coordinate value groups AC1 and AC2 of the measurement points of the end turns, and the upper inclined line is the coordinate value group L of the measurement points at the lower cone 5a of the jig 5. is there. The collection of points in the middle of the figure shows the differential value at each measurement point.

  As described with reference to FIG. 9, start point candidates IP1, IP2 and end point candidates EP1, EP2 are recognized as shown in FIG. OP1 and OP2 are excluded from processing because there are no adjacent end point candidates.

The two end points of the arc coordinate value groups AC1 and AC2 are identified by the start point candidates IP1 and IP2 and the end point candidates EP1 and EP2 recognized in this way as shown in FIG. 10, and the arc coordinate value AC1 is determined by the coordinate values continuous between the end points. , AC2 is extracted as data.
[Shape measurement program and method]
(Overall processing)
FIG. 11 is a flowchart of calculation processing in shape measurement.

  When the processing of the data in the arithmetic unit 13 is started, the preprocessing of FIG. 11 and the extraction of the wire portion are sequentially performed, and the winding shape data is output through regression calculation and the like.

  In step S1 (hereinafter, each step is abbreviated as S), the raw coordinate data of the measurement point as shown in FIG. 6 is read. The measurement point raw data is a coordinate value group of measurement points measured by the laser displacement sensor 7 as described above.

  In the preprocessing, the above-described functions are executed in steps S2, S3, and S4 with S1 as an input.

  In the “abnormal value removal” process of S2, the abnormal value is removed.

  In the “vertical coordinate sorting” process in S3, data is sorted with respect to vertical coordinates.

  In the processing of “aggregation of overlapping vertical coordinate data” in S4, when there are a plurality of displacement data at the same position, the average value is set as the displacement at that position.

  When the preprocessing is completed, the “measurement point data” in S5 as shown in FIG. 6 is output, and the process proceeds to the wire portion extraction process with S5 as an input.

  In the extraction of the wire portion, the above-described function is executed in steps S6, S7, S8, and S9.

  In the “smoothing” process of S6, the measurement point data is smoothed as described above.

  In the process of “Numerical differential calculation” in S7, the inclination of the measurement point is obtained by numerical differentiation as described above.

  In the process of “start point / end point candidate extraction” in S8, the differential values are grouped as described above to extract start point candidates and end point candidates. Details of this processing will be described later with reference to the subroutines of FIGS.

  In the process of “extracting the arc portion” in S9, the closest ones are extracted so that the start point candidate and end point candidate extracted in S8 are paired, and the position is output as arc configuration information. Details of this processing will be described later with reference to a subroutine shown in FIG.

  By extracting the arc configuration information, the wire portion extraction process ends, and the process proceeds to a calculation process of the cross-sectional center position and coil shape of the wire.

  In the calculation processing of the cross-sectional center position of the wire and the coil shape, steps S10, S11, S12, S13, S14, S15, and S16 are executed.

  In S10, the arc configuration information extracted in S9 is read by the process of “arc configuration information”.

  In the “arc point group extraction” process of S11, the “measurement point data” read in S5 and the “arc configuration information” read in S10 are input, and the arc configuration information is used from the measurement point data. Thus, a point group constituting each arc is extracted.

  In the process of “arc constituent point group” in S12, the arc constituent point group extracted in S11 is read. The “arc constituent point group” represents a point group that constitutes each arc.

  In the “regression calculation” process in S13, a circle regressed as shown in FIG. 7 by the above-described circular regression with the arc composing point group read in S12, that is, the center coordinates and the radius are obtained.

  In the process of “removing arc with large error” in S14, if the error is large by comparing with the regression error of the circle that has been regressed or the nominal wire diameter, it is removed.

  In the process of “line cross section position data” in S15, the center position and radius of each circle other than those removed in S14 are obtained and used as line cross section position data.

  In the “sort by number of turns” process of S <b> 16, the line section position data is arranged along the number of turns of the coil spring 3.

  In the “rolling shape data” process of S17, the line cross-section position data arranged along the number of turns as a measurement result is output as winding shape data. The winding shape data is output as the center position of the circular cross section of the wire arranged in a coil shape, and is displayed, for example, on the display unit 15 (FIG. 1).

(Start point candidate extraction process)
FIG. 12 shows an example of a start point candidate extraction subroutine, in which the start point candidate processing of “start point / end point candidate extraction” in S8 is performed. Here, the input to the subroutine is prepared as an array arranged in ascending order of the z-axis coordinates in the “differential value array”. “Derivative value array length” represents the number of data in this array. “Within group” is a logical variable, “Maximum differential value within group” is a real type variable, “Maximum differential value number within group” and “i” are integer type variables, and are temporary variables for processing. is there. “: =” Indicates that a value obtained by evaluating the expression on the right side is assigned to the variable on the left side.

  In the process of “maximum differential value in group: = 0, in group: = false” in S801A, initial setting is performed. “False” indicating that it is outside the group is assigned to “inside group”, and “maximum differential value within group” is initialized.

  The first differential value is specified by the process of “i: = 1” in S802A, and the process proceeds to S803.

  In the comparison process of S803A, it is determined whether or not the i-th differential value to be processed is equal to or smaller than the differential value array length. If the i-th differential value is less than or equal to the differential value array length (i <= differential value array length), the process proceeds to S804A, and if the differential value exceeds the differential value array length, the process proceeds to S812A as “else”. . A loop is formed by S802A, S803A, and S811A, and S804A to S810A are iteratively processed for all values of the “differential value array”.

  In the comparison process of S804A, it is determined whether or not the i-th differential value to be processed is within the grouping range. If the i-th differential value is less than or equal to the maximum threshold value (maximum threshold value> = differential value array (i)) and exceeds the increase / decrease threshold value (differential value array (i)> increase / decrease threshold value), the process proceeds to S805A as being within the grouping range. Otherwise, the process proceeds to S808A as “else”.

  In the process of “inside group: = true” in S805A, since the condition within the grouping range in S804A is satisfied, “true” indicating that it is within the group is substituted for the processed i-th differential value. , The process proceeds to S806A.

  In the comparison processing in S806A, it is determined whether or not the i-th differential value exceeds the maximum differential value up to the i-th in the current group. If the i-th differential value exceeds the “maximum differential value in group”, the process proceeds to S807A, and if not, the process proceeds to S811A.

  The processed i-th differential value is updated and stored as the maximum differential value in the group by the processing of “maximum differential value in group: = differential value array (i)”, maximum differential value number in group: = i ”of S807A. To do.

  The next differential value is specified by the process of “i: = i + 1” in S811A, and the process proceeds to S803A.

  The processing of S808A to S810A is a case where the i-th differential value is outside the group.

  In the comparison processing in S808A, it is determined whether or not the previous differential value is within the group and the intra-group maximum differential value exceeds the minimum threshold value. If the previous differential value is in the group and the maximum differential value in the group exceeds the minimum threshold value (maximum differential value in the group> minimum threshold value), the process proceeds to S809A, and if not, the process proceeds to S810A as “else”. .

  In the process of “output maximum intra-group differential value number as start point candidate” in S809A, the start point candidate is output and stored. The previous differential value group forms one group, for example, G1 in FIG. 9, and the group is determined. Since the maximum differential value OP1 in the group G1 exceeds the minimum threshold value, the number indicating the position Is output as a starting point candidate and stored.

  In the processing of “inside group: = false, maximum in group differential value = 0” in S810A, “false” is assigned to “inside group” because the current differential value is outside the group, and the next group The maximum differential value in the group is initialized for detection, and the process proceeds to S811A.

  The processing of S812A and S813A relates to the processing of the last group.

  In the comparison process of S812A, if the last differential value is within the group and the maximum differential value within the group exceeds the minimum threshold value (maximum differential value within group> minimum threshold value), the process proceeds to S813A, otherwise “else” is set. The starting point candidate extraction process ends.

  In the process of “output the largest differential value number in the group as a starting point candidate” in S813A, the starting point candidate is output and stored. The differential value group up to the last point of the “differential value array” forms one group, for example, G5 in FIG. 9, and the group is determined. Since the maximum differential value in the group G5 exceeds the minimum threshold value, The number indicating the position is output and stored as a starting point candidate, and the starting point candidate extraction process is terminated.

(End point candidate extraction process)
FIG. 13 shows an example of an end point candidate extraction subroutine, which performs end point candidate processing of “start point / end point candidate extraction” in S8.

  Each step in FIG. 13 corresponds to the start point candidate extraction subroutine in FIG. 12 on a 1: 1 basis. In FIG. 13, A of each step number in FIG.

  In the extraction of the end point candidate, since the sign of the differential value is positive / negative, in FIG. 13, the threshold value is given as “−”, “maximum” in FIG. 12 is “minimum” in FIG. 13, and the inequality sign in FIG. In FIG. 13, the inequality sign is reversed with respect to the direction.

  Therefore, the end point candidate can be output in the same manner as in FIG. 12 by the end point candidate extraction subroutine of FIG.

(Arc candidate extraction processing)
FIG. 14 shows an example of the arc candidate extraction subroutine, in which the “arc part extraction” process of S9 is performed.

  Initial setting is performed by “start point number = 0” in S901, and the process proceeds to S902.

  The first differential value is specified by the process of “i: = 1” in S902, and the process proceeds to S903.

  In the comparison processing in S903, it is determined whether or not the i-th differential value to be processed is equal to or shorter than the differential value array length. If the i-th differential value is less than or equal to the differential value array length (i <= differential value array length), the process proceeds to S904. If the differential value exceeds the differential value array length, there is no differential value and “else” is extracted as an arc candidate. The process ends. A loop is formed by S902, S903, and S911, and S904 to S910 are iteratively processed for all i indicating the position of the “differential value array”.

  In the comparison processing of S904, it is determined whether the i-th is a starting point candidate based on the stored data. If the i-th candidate is the starting point candidate, the process proceeds to S905, and if it is not the starting point candidate, the process proceeds to S906.

  By the process of “start point number: = i” in S905, the i-th differential value is stored as the start point candidate number, and the process proceeds to S906.

  In the comparison process of S906, it is determined whether the i-th is an end point candidate based on the stored data. If the i-th is an end point candidate, the process proceeds to S907, and if it is not an end point candidate, the process proceeds to S911.

  By the processing of “end point number: = i” in S907, the i-th differential value is stored as the end point candidate number, and the process proceeds to S908.

  In the comparison processing in S908, if the start point number is positive, that is, a valid number that is not 0, and the chord length between the start point and the end point is within the threshold, the process proceeds to S909 because there is a corresponding closest start point candidate. Otherwise, the process proceeds to S911 as “else”.

  By the processing of “output start point number to end point number as arc candidate” in S909, the start point number and end point number are paired and this number is output as an arc candidate and stored.

  In the process of “start point number = 0” in S910, the start point number is initialized to extract the next arc candidate, and the process proceeds to S911.

  The next differential value is specified by the process of “i: = i + 1” in S911, and the process returns to S903. Thereafter, the processes of S903 to S911 are repeated.

  The above subroutine is an example, and is described as three separate loops. However, all of the subroutines can be processed in one loop.

  By the above processing, as shown in FIG. 8, the continuous arc of the closely wound portion can be accurately separated.

  In addition, image processing is not required, and processing can be performed easily and reliably.

  In addition, since the differential value of each coordinate value is used, it is not necessary to know the both end points of the arc as a processing reference, the processing is simple, and a linear coordinate value group such as a jig is detected. Can be processed as is.

[Shape measuring device]
FIG. 15 is a schematic configuration diagram of a shape measuring apparatus according to the second embodiment. In addition, about a present Example, the same code | symbol is attached | subjected to the same component as Example 1, A is attached to the same code | symbol to the corresponding component, and the overlapping description is abbreviate | omitted.

  As shown in FIG. 15, in the shape measuring apparatus 1 </ b> A of the present embodiment, the laser displacement sensor 7 is arranged inside the coil spring 3. The laser displacement sensor 7 is attached so as to be capable of turning around an axis and moving up and down along the axial direction.

  In order to enable the above-mentioned turning and raising / lowering operations, in the movable portion 9A, the vertical drive mechanism 25 is disposed on the shaft core portion of the coil spring 3, and the lifting portion 19 can be moved up and down along the shaft core portion of the coil spring 3. ing. A laser displacement sensor 7 is attached to the elevating part 19, and can be moved up and down along the axial core part of the coil spring 13 as the elevating part 19 moves up and down.

  The vertical drive mechanism 25 is supported and fixed on the rotary table 23A. The rotary table 23A is attached to a table rotary motor 27A. When the rotating table 23A is rotated by driving the table rotating motor 27A, the laser displacement sensor 7 is turned around the spring axis inside the coil spring together with the movable portion 9A of the vertical driving mechanism 25.

  The laser optical axis of the laser displacement sensor 7 is supported so as to be directed radially outward from the turning center, and the same measurement as described above can be performed on the inner diameter side of the coil.

  A jig 5A for holding the coil spring 3 is provided on the fixed base 21A. A through hole 211 through which the vertical drive mechanism 25 passes is provided in the fixed base 21A.

  The jig 5A includes a lower mounting jig 5a disposed on the fixed base 21A, an upper mounting jig 5c disposed so as to face the lower mounting jig 5a from above, and a lower mounting jig 5a, It is comprised by the attachment shaft 5b which supports the upper attachment jig | tool 5c.

  The lower mounting jig 5a includes a tapered hole-shaped recess 51a and a through-hole 52a that communicates with the recess 51a and is coaxial with the through-hole 211.

  The upper mounting jig 5c is formed in the same shape as the lower mounting jig 5a, has a recess 51c, and a through hole 52a communicating with the recess 51c, and is disposed downward so as to face the lower mounting jig 5a. ing.

  The lower mounting jig 5a and the upper mounting jig 5c are set so as to be fitted and guided to the mounting shaft 5b erected on the fixed base 21A. The number of mounting shafts 5b is not limited, but three or four are set to stably guide the fitting of the lower mounting jig 5a and the upper mounting jig 5c.

  When fixing the coil spring 3 with the jig 5A, first, with the upper mounting jig 5c removed, the coil spring 3 as a shape measurement object is inserted so that the vertical drive mechanism 25 is on the axis side, and the lower mounting is performed. It arrange | positions on the jig | tool 5a. After the coil spring 3 is inserted until the lower end of the coil spring 3 comes into contact with the inner peripheral surface of the concave portion 51a of the lower mounting jig 5a, the upper mounting jig 5c is fitted and guided to the mounting shaft 5b to obtain the state shown in FIG.

  In this state, the upper mounting jig 5c is fixed to the mounting shaft 5b and positioned in a state where the coil spring hardly bends, or is positioned in a state where the upper weight of the upper mounting jig 5c works and the coil spring hardly bends. . By this positioning, the coil spring 3 is centered between the tapered hole-shaped recesses 51a and 51c, and the fixing of the coil spring 3 by the jig 5A is completed.

  Therefore, in the present embodiment, the laser displacement sensor 7 irradiates laser light from one side of the cross-sectional outer shape, that is, from the inside of the coil spring 3 to receive the reflected light from the inner portion of the coil spring 3, and the same as in the first embodiment. The effect of this can be achieved.

Further, in this embodiment, since the turning and raising / lowering drive mechanism of the laser displacement sensor 7 is disposed in the coil spring 3, the structure is advantageous in terms of space.
[Others]
In the first and second embodiments, the laser displacement sensor 7 side is rotated around or inside the fixed coil spring 3, but the laser displacement sensor 7 is fixed and the coil spring 3 is rotated. You can also

  In the first and second embodiments, the circular arc shape along the cross section of the wire 3a of the coil spring 3 is measured to identify the cross-sectional center position. However, the present invention can be similarly applied to other objects such as clips. A single cross-sectional center position may be detected.

  In the first and second embodiments, the present invention is applied to a three-dimensional coil spring. However, the present invention can also be applied to a bent pipe or a flat coil.

1, 1A shape measuring device 3 coil spring (object)
7 Laser displacement sensor 11 Control unit 13 Calculation unit 15 Display unit

Claims (12)

A shape measuring device having a function of measuring the shape of an object having an arc in a cross-sectional outer shape,
A laser displacement sensor that detects a coordinate value of a measurement point that irradiates and reflects laser light from one side of the cross-sectional outer shape at a plurality of measurement points along one side of the cross-sectional outer shape;
An arithmetic unit that approximates the shape of the arc from the coordinate value group of the plurality of measurement points,
The calculation unit obtains a differential value at each coordinate value and compares them with each other, and determines both end points of the coordinate value group corresponding to the arc based on the determination of the maximum value / minimum value of the differential value of the coordinate value group corresponding to the arc. Obtaining the approximation from the coordinate values continuous between the two end points,
A shape measuring apparatus characterized by that. A shape measuring device having a function of measuring the shape of an object having an arc in a cross-sectional outer shape,
A laser displacement sensor that detects a coordinate value of a measurement point that irradiates and reflects laser light from one side of the cross-sectional outer shape at a plurality of measurement points along one side of the cross-sectional outer shape;
A calculation unit that approximates the arc from the coordinate value group of the plurality of measurement points and calculates the center position thereof,
The calculation unit obtains a differential value at each coordinate value and compares them with each other and based on the maximum / minimum value determination of the differential value of the coordinate value group corresponding to the arc, both end points of the coordinate value group corresponding to the arc And the approximation is performed from coordinate values continuous between the two end points.
A shape measuring apparatus characterized by that. The shape measuring device according to claim 1 or 2,
The coordinate value group corresponding to the arc is a series of coordinate values in which the differential values are continuous.
A shape measuring apparatus characterized by that. The shape measuring device according to any one of claims 1 to 3,
The arithmetic unit ignores the one end point when there is no other end point candidate constituting both end points of the coordinate value group corresponding to the arc with respect to the obtained one end point,
A shape measuring apparatus characterized by that. The shape measuring device according to any one of claims 1 to 4 ,
The object has a plurality of the arcs,
The laser displacement sensor detects the coordinate value in each arc;
The calculation unit obtains the end point for each group obtained by grouping the coordinate value group corresponding to each arc according to each arc with a threshold value set for the differential value .
A shape measuring apparatus characterized by that. The shape measuring apparatus according to claim 5 ,
The object is a coil spring formed of a wire having a circular cross section,
The laser displacement sensor detects a plurality of groups of coordinate values of the measurement points along the wire rod of the coil spring,
The calculation unit approximates the shape of the circular section in each of the plurality of coordinate value groups, calculates a center position of the circular section along the wire, and obtains a coil shape of the coil spring.
A shape measuring apparatus characterized by that. A shape measurement program for causing a computer to realize a function of measuring the shape of an object having an arc in a cross-sectional outer shape,
The function is that the shape of the arc is determined from a group of coordinate values detected at a plurality of measurement points along one side of the cross-sectional outline shape by irradiating laser light from one side of the cross-sectional outline shape and reflecting it. It has a calculation function that approximates
The calculation function obtains a differential value at each coordinate value and compares them with each other, and based on the maximum value / minimum value determination of the differential value of the coordinate value group corresponding to the arc, both end points of the coordinate value group corresponding to the arc And the approximation is performed from coordinate values continuous between the two end points.
A shape measurement program characterized by that. A shape measurement program for causing a computer to realize a function of measuring the shape of an object having an arc in a cross-sectional outer shape,
The function is that the shape of the arc is determined from a group of coordinate values detected at a plurality of measurement points along one side of the cross-sectional outline shape by irradiating laser light from one side of the cross-sectional outline shape and reflecting it. With a calculation function to calculate the center position
The calculation function obtains a differential value at each coordinate value and compares them with each other, and based on the maximum value / minimum value determination of the differential value of the coordinate value group corresponding to the arc, both end points of the coordinate value group corresponding to the arc And the approximation is performed from coordinate values continuous between the two end points.
A shape measurement program characterized by that. A computer-readable shape measurement program recording medium recording a shape measurement program for causing a computer to realize a function of measuring a shape of an object having an arc in a cross-sectional outer shape,
The function is that the shape of the arc is determined from a group of coordinate values detected at a plurality of measurement points along one side of the cross-sectional outline shape by irradiating laser light from one side of the cross-sectional outline shape and reflecting it. It has a calculation function that approximates
The calculation function obtains a differential value at each coordinate value and compares them with each other, and based on the maximum value / minimum value determination of the differential value of the coordinate value group corresponding to the arc, both end points of the coordinate value group corresponding to the arc And the approximation is performed from coordinate values continuous between the two end points.
A shape measurement program recording medium characterized by the above. A computer-readable shape measurement program recording medium recording a shape measurement program for causing a computer to realize a function of measuring a shape of an object having an arc in a cross-sectional outer shape,
The function is that the shape of the arc is determined from a group of coordinate values detected at a plurality of measurement points along one side of the cross-sectional outline shape by irradiating laser light from one side of the cross-sectional outline shape and reflecting it. With a calculation function to calculate the center position
The calculation function obtains a differential value at each coordinate value and compares them with each other, and based on the maximum value / minimum value determination of the differential value of the coordinate value group corresponding to the arc, both end points of the coordinate value group corresponding to the arc And the approximation is performed from coordinate values continuous between the two end points.
A shape-measuring program recording medium readable by a computer. A shape measuring method for measuring the shape of an object having an arc in a cross-sectional outer shape,
A detection step of detecting coordinate values of measurement points reflected and reflected by laser light from one side of the cross-sectional outer shape by a laser displacement sensor at a plurality of measurement points along one side of the cross-sectional outer shape;
A calculation step of approximating the shape of the arc from the coordinate value group of the plurality of measurement points,
In the calculation step, a differential value is obtained with each coordinate value and compared with each other, and the end points of the coordinate value group corresponding to the arc are determined based on the maximum / minimum value determination of the differential value of the coordinate value group corresponding to the arc. Obtaining the approximation from the coordinate values continuous between the two end points,
A shape measuring method characterized by the above. A shape measuring method for measuring the shape of an object having an arc in a cross-sectional outer shape,
A detection step of detecting coordinate values of measurement points reflected and reflected by laser light from one side of the cross-sectional outer shape by a laser displacement sensor at a plurality of measurement points along one side of the cross-sectional outer shape;
A calculation step of approximating the shape of the arc from the coordinate value group of the plurality of measurement points and calculating the center position thereof,
The calculation step finds differential values at the respective coordinate values, compares them with each other, and determines both end points of the coordinate value group corresponding to the arc based on the determination of the maximum value / minimum value of the differential value of the coordinate value group corresponding to the arc. And the approximation is performed from coordinate values continuous between the two end points.

Images