JP5204575B2 - Position measuring method and position measuring apparatus - Google Patents

Description

  The present invention relates to a method and apparatus for measuring the position of an object.

  At automobile production sites, a sealer for rust, vibration, and soundproofing is applied to a car body that has been transported to a predetermined location and stopped. Usually, this painting operation is automated by using a sealing gun mounted on the robot arm, and the drive control of the robot arm is executed so that the sealer discharged from the sealing gun is applied to a predetermined part of the vehicle body.

  The precision of the vehicle stop position is lower than the accuracy required for the painting work because precise initial control and maintenance costs are required if the control of the transport and stop of the car body is performed precisely in order to apply the sealant to the appropriate part. It tends to be. For this reason, the apparatus for controlling the robot arm first measures the position of the reference point provided on the stopped vehicle body, and calculates a correction value corresponding to the measured position deviation amount from the master data stored in advance. . The drive control of the robot arm is executed based on the position data obtained by correcting the master data using this correction value, so that the sealer agent can be applied to an appropriate part without excessively increasing the accuracy of the stop position of the vehicle body. (For example, refer patent document 1).

Patent Document 1 is configured to use desk-top design data such as CAD values and CAM values as master data, but correct the master data according to the measured three-dimensional position data. A position measuring device is disclosed. In addition, this position measuring apparatus is configured such that one camera is provided for each of the six robot arms. These six cameras are divided into three groups, and two cameras in each group capture the reference point by stereo vision. Based on the imaging result, three-dimensional position data of the three reference points is acquired.
Japanese Patent No. 2767417

  However, in practice, it is difficult to produce a car body made by welding a plurality of steel plates according to the design data, so it is unavoidable between the dimensions of the car body actually produced and the dimensions of the design data. There is a deviation. For this reason, when the design data is used as master data as in Patent Document 1, the deviation between the actual dimension and the designed dimension remains in the corrected data. In this way, when design data is used as master data, it is difficult to accurately calculate a correction value due to the generation of a unique solution, and the application position of the sealer may be shifted.

  In addition, according to Patent Document 1, one camera is provided for each robot arm, and a stereo image is generated by two cameras attached to separate robot arms. Since the robot arm inevitably bends downward in the vertical direction based on the weight of the camera and its own weight, the deflection of each robot arm causes the stereo base and the included angle to deviate from the desired values, thereby increasing the position measurement of the reference point. It has become difficult to perform accurately. If the robot arm drive control is performed in consideration of the bending of the robot arm in advance, the control content becomes complicated and the time required for setting the camera becomes longer. For this reason, the time required for the painting work becomes longer, which may affect the productivity of the automobile.

  Therefore, an object of the present invention is to accurately measure the position of an object and further to accurately correct position data of the object.

The position measurement method according to the present invention calculates a correction value for master data serving as a reference for the position of the object based on position data obtained by actually measuring the object, and a correction position in which the correction value is added to the master data. A position measurement method applied to an apparatus configured to work on a predetermined part of the object based on data, wherein the position of a plurality of reference points provided in each of a plurality of samples of the object is measured. A pre-measurement step for acquiring position data of each reference point, and the master data for each of the plurality of reference points in advance based on the position data of the plurality of reference points of each sample acquired in the pre-measurement step and pre-calculating step of calculating, the measured steps of measuring the positions of a plurality of reference points provided on the object to obtain position data of each reference point, obtained in said actual measuring step Have a correction value calculation step of calculating a correction value for the master data calculated by the pre-calculation step on the basis of the position data, the correction value calculation step, the positional relationship between the master data, each Defining a first circle through which the master data passes, defining a positional relationship between each reference point measured in the actual measurement step with a second circle through which each reference point passes, and the first circle Calculating a coordinate conversion amount for converting the first coordinate system that defines the second circle into the second coordinate system that defines the second circle, and the expression L1 (φ) ² + L2 (φ) ² +. + Ln (φ) ² to obtain an angle φ when ² takes a minimum value, where φ is a rotation angle when rotating around the normal of the first circle, and Ln (φ) is The first circle is normal times by an angle φ Is the distance between the n-th master data and the position data of the reference point corresponding to the n-th master data, and the coordinate transformation amount and the angle φ are the correction values. It is characterized in that have a step of calculating a.

According to this configuration, master data serving as a reference for the position of the reference point of the object is created using the sample of the object, and a correction value for correcting the master data based on the actually measured position data is provided. Calculated. In this way, ideal design data is not used as master data, but data obtained by actual measurement is used as master data, and a correction value is calculated based on position data that is also actually measured. Work is performed on the object based on the data obtained by correcting the data. For this reason, the operation | work with respect to a target object can be performed correctly. In addition, the correction value is calculated using the least square method, and correction can be performed with a good balance for each of the plurality of reference points.

  In the preliminary measurement step, the three-dimensional position data is acquired based on a pair of images obtained by photographing the sample by stereo vision. In the actual measurement step, the pair of images obtained by photographing the object by stereo vision is obtained. Based on this, the three-dimensional position data may be acquired. According to this configuration, three-dimensional position data is acquired, and advanced position measurement is realized.

  In the pre-calculation step, the master data may be calculated by averaging position data of reference points of each sample. According to such a configuration, it is possible to easily calculate master data, and it is possible to create highly reliable data by increasing the number of samples.

A position measuring device according to the present invention includes a storage unit that stores master data that serves as a reference for the position of a measurement object, and a measurement unit that measures position of a plurality of reference points provided on the object to obtain position data. And a correction value calculation unit for correcting the master data based on the position data acquired by the measurement unit, the measurement unit by the measurement unit prior to actual measurement of the measurement object The position of a plurality of reference points provided in each of the plurality of samples of the measurement object is measured, and the position data of the plurality of reference points of each sample is acquired. Based on the acquired position data in the storage unit The calculated master data is stored, and at the time of actual measurement of the measurement object, the correction value calculation unit converts the master data stored in the storage unit into the position data of the reference point of the measured measurement object. Ri configuration der for correcting Zui, the correction value calculation unit, a positional relationship between said master data, said defined by a first circle which each master data passes, the positional relationship between the reference points is measured Defined by a second circle through which each reference point passes, and calculates a coordinate conversion amount for converting the first coordinate system defining the first circle to the second coordinate system defining the second circle Then, the angle L when the expression L1 (φ) ² + L2 (φ) ² +... + Ln (φ) ² takes the minimum value is obtained, and the coordinate transformation amount and the angle φ are calculated as correction values. , Φ is the rotation angle when rotating around the normal of the first circle, and Ln (φ) is the nth master when the first circle is rotated around the normal by the angle φ the distance der Rukoto between data and said of n-th master data and the position data of the corresponding reference points characterized ing. Even in such a device, the same operation as described above occurs.

  First and second cameras for photographing the reference point by stereo vision are provided, and the measurement unit measures position data of the reference point based on a pair of images photographed by the first and second cameras. Preferably, the first and second cameras are attached to a single robot arm. According to such a configuration, the stereo base and the included angle are not changed even if the robot arm is bent by attaching the stereo camera to the same robot arm. Therefore, position measurement based on an image captured by stereo vision can be performed with high accuracy.

  The position measuring device according to claim 7, wherein a working device that performs a predetermined work on an object based on the position data corrected by the correction unit is mounted on the robot arm. According to such a configuration, it is possible to shift to a predetermined operation immediately after the measurement of the position based on the imaging result and the correction process of the position data is completed, and it is possible to shorten the time required for the operation using the position measurement. .

  As described above, according to the present invention, it is possible to accurately correct the position data of an object without generating a singular solution.

  Embodiments according to the present invention will be described below with reference to the accompanying drawings.

  Here, an example of an object according to the present invention will be described as a vehicle body. FIG. 1 is an explanatory diagram schematically showing a partial outline of a vehicle production site in a plan view, and FIG. 2 is an explanatory diagram schematically showing it in a rear view. The production site shown in FIGS. 1 and 2 is provided with a rail 1 that forms a conveyance path for a vehicle body 100 formed by welding a large number of steel plates. A plurality of hangers 3 are suspended from the rail 1 via a traveling unit 2, and each hanger 3 is engaged with a window frame portion 101 on the side of the vehicle body 100. The vehicle body 100 suspended by the hanger 3 in this way is conveyed along the rail 1 based on the operation of the traveling unit 2.

  In this conveyance path, a coating area 5 for sealing and spray-coating a sealant for rust prevention, soundproofing, and vibration proofing is set at the joint of the floor portion 102 of the vehicle body 100 and the steel plate. Yes. In the painting area 5, a plurality of articulated robots 6a, 6b, 6c, 6d are provided. Each robot 6a, 6b, 6c, 6d has a plurality of arms 7 and a joint 8 that connects the arms so as to be swingable. The tip arm 9 of the robot 6a, 6b, 6c is used to measure the position of a sealing gun 25 used to apply a rust-proof and waterproof sealer material to the vehicle body, which will be described later, and the position of the vehicle body that has been transported and stopped. A camera 33 and the like are mounted, and a sealing gun 25 is mounted on the robot 6d. FIG. 3 is a left side view for referring to the overall appearance of the robots 6a, 6b, and 6c. The tip arm 9 has one sealing gun 25 and two cameras 33 arranged side by side. Is installed.

  The painting area 5 has a cut shape, and the vehicle body 100 is conveyed between the pair of upper floor surfaces 10. Each of the upper floor surfaces 10, 10 and the lower floor surface 11 is provided with slide rails 12 extending in the conveyance direction of the vehicle body 100. The robots 6a, 6b, and 6c are slidable along the slide rail 12 provided on the upper floor surface, and the robot 6d is slidable along the slide rail 12 provided on the lower floor surface.

  Further, the painting area 5 is provided with a calibration stand 13 which is used for confirming an origin position which will be described later and in which a calibration hole 13a is formed.

  4 is a front perspective view of the tip arm 9, FIG. 5 is a rear perspective view thereof, FIG. 6 is a front view thereof, FIG. 7 is a rear view thereof, FIG. 8 is a plan view thereof, FIG. FIG. 11 is a left side view thereof, and FIG. 12 is a front sectional view thereof. The direction of the following robot hand 9 is such that when the optical axis of the camera is oriented horizontally, the subject side of the camera is the front side and the opposite side is the rear side for convenience. However, the concept of this direction is appropriately changed according to the operation of the robot in the painting area.

  As shown in FIGS. 4 to 12, the distal arm 9 has a base bracket 21 swingably attached to the joint 8, and a bracket 23 extending to the distal end side is fastened to the base bracket 21.

  A stay 24 extending further toward the distal end side is fastened to the distal end surface of the bracket 23. An inclined surface 24a that is inclined with respect to the extending direction of the bracket 23 is formed at the tip of the stay 24, and a sealing gun 25 is fastened to the inclined surface 24a. The sealing gun 25 is a device that injects the sealing agent onto the vehicle body 100 (see FIGS. 1, 2, and 15), and a hose 26 that supplies the sealing agent is connected to the sealing gun 25. A hose bracket 27 for fixing the hose 26 is fastened to the bracket 23 on the outside.

  A camera bracket 28 is fastened to the bracket 23. The camera bracket 28 includes an upright portion 29 provided so as to stand up from the bracket 23, and an attachment portion 30 that is fastened to the upper end surface of the upright portion 29 and extends in parallel with the bracket 23, and the upright portion 29 and the attachment portion. 30 is T-shaped in a front view.

  The attachment portion 30 extends toward the distal end side with respect to the standing portion 29. Therefore, a rib 31 that protrudes toward the tip side and is fixed to the back surface of the mounting portion 30 is joined to the tip surface of the standing portion 29, and the weight of the member fixed to the mounting portion 30 is supported by the rib 31. It has been. Thereby, bending of attachment part 30 is prevented.

  The mounting portion 30 extends in the left-right direction with respect to the standing portion 29, and one camera holder 32 is fastened to each of the left and right end portions of the upper surface of the mounting portion 30, and each camera holder 32 has a cylindrical shape. A camera 33 having a lens barrel 34 is held. The camera 33 includes a photoelectric conversion element such as a CCD or a CMOS, and is configured to convert an imaging result into an electrical signal in the photoelectric conversion element and output it to the outside. The focal length, magnification, and the like of the camera 33 are set to predetermined values appropriate for photographing with the dial 35 provided on the outer surface of the lens barrel 34. Note that a bracket extending from the rib 31 in the extending direction of the mounting portion 30 may be provided on the back surface of the mounting portion 30 in order to prevent the downward bending of the mounting portion 30 to which the camera 33 is mounted.

  A camera cover 36 is fastened to the attachment portion 30, and the entire surface of the camera 33 is covered by the camera cover 36. The pair of cameras 33 are arranged so that their optical axes intersect to enable image processing by stereo vision. As described above, the two cameras 33 that are viewed in stereo are installed on the same camera bracket 28, so that the stereo base can be used regardless of the operation of the arm 7 (see FIG. 1) and the vertical downward deflection of the arm 7. δ and the included angle θ are constant. Therefore, in a so-called hand-eye type, position measurement based on a stereo image can be performed with precision.

  A lamp cover 37 is fastened to the front end surface of the mounting portion 30 between the pair of camera covers 36, and an illumination light source 38 such as a ring fluorescent lamp is attached by the lamp cover 37.

  The hose 26 is disposed below the camera 33, the lamp cover 37 is disposed behind the objective lens of the camera 33, and the sealing gun 25 is disposed between the pair of cameras 33. For this reason, various components are fixed to the bracket 23. However, these components are not located in the field of view of the camera, and the collective arrangement of these components is realized thereon.

  FIG. 13 is a block diagram illustrating the configuration of a control system that processes the image captured by the camera 33 and controls the operation of the sealing gun 25. A control device 40 shown in FIG. 13 is a microcomputer including a CPU 41, a ROM 42, a RAM 43, and an input / output interface (hereinafter referred to as I / F) 44, and these elements 41 to 44 are connected via a bus 45. . The CPU 41 executes a program stored in the ROM 42 based on information input via the I / F 44 and information stored in the RAM 43, and controls operation of devices connected via the I / F 44.

  Six cameras 33 are connected to the I / F 44 as input devices. These cameras 33 are divided into three groups, and each group includes two cameras 33 that are viewed in stereo. The two cameras 33 constituting each set are attached to the same camera bracket 28 (tip arm 9) as described above. For this reason, as described above, position measurement based on an image obtained by stereo viewing can be performed with precision by the two cameras 33 in each group.

  A plurality of the sealing guns 25, a plurality of travel shaft motors 46, and a plurality of arm drive motors 47 are connected to the I / F 44 as output devices. The travel axis motor 46 is a drive source that causes each robot 6a, 6b, 6c, 6d (see FIGS. 1 and 2) to travel along the slide rail 12 (see FIGS. 1 and 2). The arm drive motor 28 is a drive source that drives the joint 8 of each robot 6a, 6b, 6c, 6d to swing the arm 7 (see FIGS. 1 and 2). In addition, although the form which controlled the operation | movement of the several robot 6a, 6b, 6c, 6d by the single control apparatus was illustrated here, a separate controller is distributed with respect to each robot 6a, 6b, 6c, 6d. The system configuration may be provided.

  Hereinafter, the operation of the robots 6a, 6b, 6c, and 6d and the contents of the control executed by the control device 40 will be described.

  First, a common coordinate system can be used within the movable range of each robot. Since this process itself is publicly known, its outline will be briefly described. A needle bar (not shown) is attached to each robot 6a, 6b, 6c, 6d. This needle bar is provided in place of the sealing gun 25 (see FIG. 2) while the camera 33 (see FIG. 2) remains attached. Next, an arbitrary contact point within the movement range of the robot is set. This contact point may be single or preferably plural. Moreover, it is preferable that this contact point is set to a point that is presumed that the needle bars come into contact with each other when the postures of adjacent robots are the same. This is to avoid the influence of arm deflection as much as possible.

  Next, the two robots are driven so that the tips of the needle bars attached to the two adjacent robots move to an arbitrary contact point set as described above. Then, the predetermined part of one needle bar and the predetermined part of the other needle bar are brought into point contact. The robot coordinate value and the robot coordinate value at that time are output. Based on the two coordinate values, a conversion matrix for converting the coordinate value of one robot into the other coordinate value is calculated. When a plurality of arbitrary contact points are set, this procedure is repeated for each contact point.

  By using the conversion matrix calculated in this way, the coordinates of a certain robot can be extended to the coordinate area of an adjacent robot. When a plurality of contact points are set, a conversion matrix is calculated for each contact point. For this reason, an average transformation matrix obtained by averaging the plurality of transformation matrices can also be used for the extension of coordinates. The above processing is performed for all of the adjacent robot sets.

  Next, the calibration process of the camera 33 is performed. Since this process itself is publicly known, its outline will be briefly described. First, the coordinate position of the calibration stand 13 (see FIG. 2) is measured using the needle bar. As described above, the calibration stand 13 is installed at an appropriate position in the painting area 5 and has a predetermined hole 13a. This coordinate position is measured, for example, by inserting a needle bar into the hole 13a of the calibration stand 13 and specifying the coordinate value of the robot when the needle bar is inserted into the hole 13a as the coordinate position. Thus, after measuring the coordinate position of the calibration stand 13, the camera 33 (see FIG. 2) is used to photograph the hole 13a of the calibration stand 13, and the camera dimensions are calibrated based on the imaging result of the hole 13a. Do. This calibration process is performed for all the robots 6a, 6b, and 6c that are to be equipped with a camera 33 that is used to measure a reference point, which will be described later.

  Next, before actually performing a painting operation on a certain vehicle type, the present control device 40 creates master data necessary for processing executed during the actual operation. When creating the master data, a plurality of samples (for example, 3 to 10) of the vehicle body 100 of the vehicle type that will be actually subjected to painting work in the production site are prepared, and these samples are put on the hanger 3 at the production site. Be suspended. In other words, this master data is ideal, but it is not created based on CAD values or CAM values that are design values on the desk, but for each vehicle type through pre-training at the production site where actual painting work is performed. Created one by one.

  FIG. 14 is a flowchart illustrating a procedure for creating master data. As shown in FIG. 14, first, a sample of the vehicle body 100 suspended from the hanger 3 at the production site is transported along the rail 1 and stopped in the painting area 5 (step S1). Next, a sample is imaged using the camera group 33 of the vehicle body 100 (step S2), and the position of a reference point appropriately defined in the vehicle body 100 is measured based on the image data input from the camera group 33 (step S3). .

  FIG. 15 is a conceptual diagram of the imaging process of the vehicle body 100 and the reference point position measurement process. Here, a so-called front engine type vehicle body (including a sample) 100 is illustrated. A large number of through holes are formed in the floor portion 102 of the vehicle body 100. For example, when fixing the engine to be accommodated in the front bonnet to the vehicle body 100, a small hole 103 for draining water that has entered the rear trunk. Small holes 104 and 105 used for the above are formed. Here, these small holes 103, 104, and 105 are used as reference points P 1, P 2, and P 3 that are imaging targets of the camera 33 and are targets for subsequent position measurement, respectively.

  In step S2 in FIG. 14, for example, the small holes 103 corresponding to the reference point P1 with the two cameras 33 attached to the robot 6a correspond to the reference point P2 with the two cameras 33 attached to the robot 6b. The small holes 104 corresponding to the reference point P3 are imaged in stereo, respectively, by the two cameras 33 attached to the robot 6c.

  In step S3 in FIG. 14, the small hole images 103, 104, and 105 are identified by pattern matching from the two images captured by the two cameras 33 of each set, and the principle of triangulation is used for the identified image positions. The three-dimensional positions as the reference points P1, P2, P3 are detected from the calculated formula. The positions of the reference points P1, P2, P3 are defined in the three-dimensional coordinate system. In addition, you may add other image processes, such as a binarization process, to this step S3 suitably.

  Returning to FIG. 14, when the position detection process (step S3) is completed, the detected position information is then stored in the RAM 43 (see FIG. 13) (step S4). Then, it is determined whether or not the number of samples for which position data has been acquired has reached a predetermined value (step S5). If the predetermined value has not been reached, the processing of steps S1 to S4 is performed again for the next sample. Done.

  When the position data of a predetermined number of samples is acquired, master data of the positions of the reference points P1, P2, P3 is created based on these position data (step S6). Each master data is created by averaging a plurality of position data acquired for each of the reference points P1, P2, P3. At this time, master data may be created after appropriately removing outliers. In order to select this outlier, it is also possible to use a CAD value or a CAM value. Then, the master data for the reference points P1, P2, P3 created in this way is stored in the RAM 43 (step S7).

  Next, a procedure for performing an actual painting operation performed using the master data created through the prior training will be described with reference to FIG. As shown in FIG. 16, first, the vehicle body 100 is transported and stopped in the painting area 5 in the same manner as in the pre-training (step S11), and the small holes 103, 104, 105 formed in the stopped vehicle body 100 are stereo. In order to correct the master data based on the measured data (step S12), measure the positions of the reference points P1 ′, P2 ′, P3 ′ based on the input image data (step S13). Is calculated (step S14).

  FIG. 17 is a flowchart for explaining the correction procedure, and FIG. 18 is a conceptual diagram of the correction process. As shown in FIGS. 17 and 18, first, a coordinate system for positioning the reference point P1 defined by the master data on the X axis is defined, and the reference points P1, P2, P3 are set with the origin O of the coordinate system as the center O. A circle C passing through is obtained. In addition, a coordinate system for setting the measured reference point P1 ′ on the X ′ axis is set, and a circle C ′ passing through the reference points P1 ′, P2 ′, P3 ′ is obtained with the origin of the coordinate system as the center O ′. (Step S41). That is, in step S41, the positional relationship between the three reference points P1, P2 and P3 with reference to the reference point P1, and between the three reference points P1 ', P2' and P3 'with reference to the reference point P1'. Are defined by circles defined in a two-dimensional coordinate system.

  Next, the circle C is translated so that the center of the circle C on the master data side coincides with the center O ′ of the circle C ′ on the measurement value side (step S42). Then, in order to make the plane defining the circle C after the parallel movement coincide with the plane defining the circle C ′ on the measurement value side, the plane defining the circle C is rotated (step S43). That is, in steps S42 and S43, processing for converting a coordinate system set with reference to the reference point P1 into a coordinate system set with reference to the reference point P1 ′ is executed.

When the circle C defined in the plane after the rotational movement is rotated by an angle φ around the normal line passing through the center O ′, L1 (φ) ² + L2 (φ) ² + L3 (φ) ² becomes An angle φ that takes the minimum value is obtained (step S44). Here, L1 (φ) is the distance between the reference point P1 after rotation by the angle φ and the measured value P1 ′, L2 (φ) is the distance between the reference point P2 and the measured value P2 ′, and L3 (φ ) Is the distance between the reference point P3 and the measured value P3 ′. That is, in step S44, using the principle of the least square method, the circle C that defines the positional relationship between the reference points in the master data is approximated to the circle C ′ that defines the positional relationship between the measured reference points. The process to be executed is executed.

  Then, the amount of movement of the circle C through steps S42 to S44 is calculated as a correction value (step S45).

  Returning to FIG. 16, the master data is corrected using the correction value calculated in this way, and a painting operation is performed on a predetermined portion of the vehicle body 100 based on the corrected data (correction position data) (step S15). . Referring to FIG. 15, by this painting operation, a sealer is applied to fender 106 and a plurality of sealing beads 107 are formed on floor back 102 of vehicle body 100. The sealing bead 107 is formed by the sealing gun 25 operated in the vehicle body 100 through the window frame portion 101. In this embodiment, since the camera 33 is mounted on the robots 6a, 6b, and 6c that perform indoor sealing in this way, it is possible to shift to the painting work immediately after the position measurement is finished. Further, the drive control of the robot 6d without the camera is also executed based on the corrected position data calculated using the robots 6a, 6b, 6c. In this way, the painting operation can also be performed by a system configuration in which the robot 6a, 6b, 6c on which the camera 33 is mounted is a master robot, and the robot 6d on which only a working device such as the sealing gun 25 is mounted is a slave robot.

  Based on the processing content of the CPU 41 as described above, the CPU 41 is measured by the measurement unit that measures the three-dimensional position of the reference point of the vehicle body 100 based on the image signal indicating the imaging result input from the camera 33 and the measurement unit. A correction value calculation unit that calculates a correction value for correcting the master data based on the position data of the reference point, and the arm 7 (see FIG. 1) based on the correction position data of the vehicle body corrected by the correction value calculation unit. The work control unit that performs the swing control and the swing control of the sealing gun 25 is functionally provided. In the measurement unit, the position of the reference point of the sample is measured in advance training, and the CPU 41 further calculates a master data of the reference point of the vehicle body based on the position data of the sample measured in advance by the measurement unit. It will have a data calculation part functionally.

  As described above, according to the present embodiment, the positions of the reference points P1 ′, P2 ′, P3 ′ measured based on the master data created in the prior training are corrected. For this reason, a correction value adapted to the actual production process can be obtained, and the painting operation can be performed appropriately. More specifically, the tolerance between the design data and the actually produced vehicle body data can be eliminated, and the operation of the welding robot used when forming the vehicle body 100 that generates this tolerance is reduced. In addition, it is possible to absorb the trap of the stopping operation of the traveling unit 2 on the master data side, and to reduce the correction error. Further, the deflection of the arm 7 of the robot 6a, 6b, 6c can be absorbed in the master data by the prior training, and the correction error can be reduced. Since the deflection of the arm 7 is generally larger than the resolution of the camera 33, it is particularly effective to be able to absorb this deflection for accurate position measurement.

  The robots 6a, 6b, and 6c equipped with the camera 33 are not arranged on the lower floor surface 11 where the sprayed sealer agent adheres and is easily soiled. For this reason, it is not necessary to provide the camera 33 with a shutter mechanism for protecting the adhesion of the sealer agent, and the peripheral structure of the camera and the configuration of the control system can be simplified. Moreover, since it is not necessary to secure the opening / closing time of the shutter mechanism, the time required for the painting work can be shortened. Since the illumination light source 38 is similarly removed from the lower floor surface 11, contamination of the illumination light source 38 can be prevented.

  In this way, since many parts related to imaging are omitted on the lower floor surface 11, the number of cameras and illumination light sources corresponding to a plurality of vehicle types can be reduced, and a painting dedicated robot is provided here. A large number can be arranged. For this reason, shortening of the conveyor length and the booth length and the shortening of the entire process time can be achieved, and the initial capital investment can be reduced. In addition, maintenance work at the production site can be performed at a low cost.

  In addition, when carrying out so-called mixed-flow production in which different vehicle types are handled as work objects at the same production site, master data is created by performing training prior to actually performing operations on the added vehicle types. . Thus, it is possible to easily cope with mixed flow production only by creating master data for each vehicle type.

  Furthermore, in this embodiment, since the cameras 33 are mounted on the robots 6a, 6b, and 6c, even if the position of the reference point is significantly changed between the vehicle types, the operations of the robots 6a, 6b, and 6c can be performed. Based on this, the reference point of each vehicle type can be easily within the field of view. Further, the robots 6 a to 6 d are movable along the slide rail 12. Therefore, even if mixed flow production is performed, it is not necessary to increase the number of cameras 33 and illumination light sources 38. Finally, the paint area 5 can be constructed in a compact manner, which is good from the environmental aspect.

  Next, another procedure for correcting the master data will be described. FIG. 19 is a flowchart for explaining the procedure, and FIG. 20 is a conceptual diagram of this correction processing. As shown in FIGS. 19 and 20, first, PQ coordinates are set such that the master data for the reference point P1 is the origin, the master data for the reference point P2 is on the P axis, and the reference point P3 is positioned on the PQ plane. A P′Q ′ coordinate system is set such that the measured reference point P1 ′ is the origin, the reference point P2 ′ is on the P ′ axis, and the reference point P3 ′ is positioned on the P′Q ′ plane. Set (step S141). As described above, in this embodiment, the positional relationship between the reference points P1, P2, and P3 is defined by positioning the reference point P3 in the two-dimensional coordinate system based on the reference points P1 and P2, and the reference point P3. 'Is positioned in a two-dimensional coordinate system with reference points P1' and P2 'as a reference, thereby defining the positional relationship between the reference points P1', P2 'and P3'.

  Next, the coordinates of the master data for the reference point P1 are translated on P1 ′ (step S142), the P axis is converted to coincide with the P ′ axis (step S143), and the Q axis is changed to the Q ′ axis. Conversion is performed so as to match (step S144). Then, the coordinate conversion amount in steps S142 to S143 is calculated as a correction value (step S145). That is, in steps S142 to S143, processing for approximating the positional relationship between the reference points is performed by moving the axes defining the coordinates.

  As described above, also in the present modified example, the measurement value is corrected based on the master data created in the pre-training, so that the coating operation can be appropriately performed based on the corrected position data after the correction. .

  In addition, since this correction method can be performed only with simple coordinate transformation, it is advantageous in that the correction value can be easily calculated as compared with the above embodiment. However, in this modified example, weighting is performed according to the order of steps S142 to S144, so that the correction accuracy of the reference point P3 is lower than the correction accuracy of the reference point P1. The above embodiment is advantageous over the present modification in that the weighting between the reference points is eliminated by using the least square method, and the positional relationship between the reference points can be corrected in a balanced manner. It has become.

  Although the embodiments according to the present invention have been described so far, the scope of the present invention is not limited to the above-described configuration, and can be changed as appropriate. For example, the focal length of the camera may be fixed or variable. Furthermore, the camera 33 installed on the attachment portion 30 may be configured to be movable so that the stereo base δ and the included angle θ change.

  In the present embodiment, an articulated robot is exemplified, but the form of the robot is not limited to this, and the present invention is also suitably applied to other forms of robots such as a cylindrical coordinate type robot and an orthogonal coordinate type robot. Further, the position measurement method can be appropriately applied not only to the painting work but also to other uses.

  The present invention can be widely applied in the technical field of measuring the position of an object.

It is explanatory drawing which shows typically the one part outline | summary of the production site of a motor vehicle by planar view. It is explanatory drawing which shows typically the one part outline | summary of the production site of a motor vehicle in back view. It is a right view which shows the external appearance of the whole robot. It is a front perspective view of a tip arm. It is a back perspective view of a tip arm. It is a front view of a tip arm. It is a rear view of a tip arm. It is a top view of a tip arm. It is a bottom view of a tip arm. It is a right view of a tip arm. It is a left side view of a tip arm. It is a front sectional view of a tip arm. It is a block diagram explaining the structure of the control system which performs the process of the image imaged with the camera, and drive control of a sealing gun. It is a flowchart explaining the creation procedure of master data. FIG. 17 is a conceptual diagram of imaging processing and reference point position measurement processing shown in FIGS. 14 and 16. It is a flowchart explaining the procedure of a painting operation. It is a flowchart explaining the procedure of a correction process. It is a conceptual diagram of the correction process shown in FIG. It is a flowchart explaining the procedure of the example of a change of a correction process. It is a conceptual diagram of the example of a change of the correction process shown in FIG.

Explanation of symbols

6a, 6b, 6c, 6d Robot 25 Sealing gun 28 Camera bracket 33 Camera 100 Car body (including sample)

Claims (6)

Based on the position data obtained by actually measuring the object, a correction value for the master data serving as a reference for the position of the object is calculated, and the predetermined value of the object is determined based on the correction position data obtained by adding the correction value to the master data. A position measurement method applied to an apparatus configured to work on a site,
A pre-measurement step of measuring the position of a plurality of reference points provided in each of a plurality of samples of the object to obtain position data of each reference point;
A pre-calculation step for pre-calculating the master data for each of the plurality of reference points based on the position data of the plurality of reference points of each sample acquired in the preliminary measurement step;
An actual measurement step of measuring the positions of a plurality of reference points provided on the object and obtaining position data of each reference point;
Have a correction value calculation step of calculating a correction value for the master data calculated by the pre-calculating step based on the positional data obtained by said actual measurement step,
The correction value calculating step includes:
Defining a positional relationship between the master data by a first circle through which each of the master data passes;
Defining a positional relationship between each reference point measured in the actual measurement step by a second circle through which each reference point passes;
Calculating a coordinate conversion amount for converting the first coordinate system defining the first circle into the second coordinate system defining the second circle;
Calculating the angle φ when the expression L1 (φ) ² + L2 (φ) ² +... + Ln (φ) ² takes the minimum value,
Here, φ is the rotation angle when rotating around the normal line of the first circle, and Ln (φ) is the nth rotation angle when the first circle rotates around the normal line by the angle φ. And the position data of the reference point corresponding to the n-th master data and the n-th master data,
Furthermore, the position measuring method characterized by have a step of calculating the coordinate conversion amount and the angle φ as the correction value. In the preliminary measurement step, based on a pair of images obtained by photographing the sample by stereo vision, the three-dimensional position data is acquired,
The position measurement method according to claim 1, wherein in the actual measurement step, the three-dimensional position data is acquired based on a pair of images obtained by photographing the object by stereo vision.   3. The position measuring method according to claim 1, wherein in the pre-calculation step, the master data is calculated by averaging position data of reference points of each sample. A storage unit for storing master data serving as a reference for the position of the measurement object;
A measurement unit that measures the positions of a plurality of reference points provided on the object and obtains position data;
A position measurement device having a correction value calculation unit for correcting the master data based on the position data acquired by the measurement unit,
Prior to actual measurement of the measurement object, the measurement unit measures the positions of a plurality of reference points provided on each of the plurality of samples of the measurement object, and acquires position data of the plurality of reference points of each sample. , The master data calculated based on the acquired position data is stored in the storage unit,
When measured of the measuring object, the correction value calculation unit, the master data stored in the storage unit, Ri configuration der to correct based on the position data of the reference points of the actually measured measurement object,
The correction value calculation unit defines a positional relationship between the master data by a first circle through which the master data passes, and a second circle through which each reference point passes through a measured positional relationship between the reference points. And a coordinate transformation amount for transforming the first coordinate system defining the first circle into the second coordinate system defining the second circle is calculated, and the expression L1 (φ) ² + L2 (Φ) ² +... + Ln (φ) ² is obtained as an angle φ when the minimum value is obtained, and the coordinate transformation amount and the angle φ are calculated as correction values,
Here, φ is the rotation angle when rotating around the normal line of the first circle, and Ln (φ) is the nth rotation angle when the first circle rotates around the normal line by the angle φ. position measuring device according to claim distance der Rukoto between the master data and said n-th master data and the position data of the corresponding reference point. A first and second camera that captures the reference point by stereo vision, and the measurement unit acquires position data of the reference point based on a pair of images captured by the first and second cameras And
The position measuring apparatus according to claim 4 , wherein the first and second cameras are attached to a single robot arm. The position measuring device according to claim 5 , wherein a working device for performing a predetermined work on the measurement object based on the position data corrected by the correcting unit is mounted on the robot arm.

Images