Elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device and feeding riveting method thereof

Abstract

The invention discloses a continuous automatic feeding and intelligent positioning riveting device for elevator guide rail gaskets, comprising a base, a rotating feeding mechanism, a bidirectional positioning feeding mechanism and a three-axis visual riveting mechanism arranged on the base; a riveting gun, a CCD camera electrically connected to a control center and a light source are fixed on the three-axis visual riveting mechanism. The invention also discloses a feeding riveting method of the device.

Description

Technical Field

[0001] The invention relates to a continuous automatic feeding and intelligent positioning riveting device for elevator guide rail gaskets. Background Art

[0002] The assembly process of T-shaped hollow elevator guide rails includes the riveting process of elevator guide rails and guide rail gaskets. It is necessary to take the guide rail gaskets (plate-shaped workpieces) out of the material area, and then send the gaskets to the workpiece positioning mechanism, send the gaskets into the interior of the guide rails through the positioning mechanism, and finally fix the gaskets to the end of the guide rails by riveting. The existing elevator guide rails and gaskets equipment, as described in patent CN201821752418.8, the guide rail gaskets are arranged on the material tray, and the gaskets are grabbed by the loading robot and then passed to the positioning mechanism, and the empty material tray is sent away by the conveying device and replaced with a new material box. The device is large in size, and the robot takes a long time to pick up the materials, which delays the production cycle of the guide rail and gasket installation.

[0003] In order to ensure the installation accuracy of the gasket, during the gasket positioning process (sending the gasket to the riveting position), the gasket needs to be constrained and positioned in two degrees of freedom, horizontally and vertically. After the gasket positioning and riveting are completed, the positioning device must withdraw from the inside of the guide rail. Since the lateral dimension of the inner wall of the guide rail is only 1 to 2 mm larger than the lateral dimension of the gasket, the conventional positioning device cannot perform bidirectional positioning of the gasket in a narrow space. In addition, if the longitudinal positioning structure of the gasket is in a fixed form, then after the gasket is riveted, the gasket is fixed to the guide rail and the positioning device cannot withdraw. In addition, the conventional feeding positioning method cannot ensure the precise alignment of the riveting holes of the gasket and the guide rail. Visual processing can be used to identify the workpiece hole position for alignment, but the assembly site environment is harsh, the workpiece surface is dusty, and the image noise points collected by the CCD camera are many and complex. It is difficult to identify the hole position using conventional image processing methods. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a continuous automatic feeding and intelligent positioning riveting device for elevator guide rail gaskets.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: an elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device, comprising a base, and a rotating feeding mechanism, a bidirectional positioning feeding mechanism and a three-axis visual riveting mechanism arranged on the base;

[0006] The rotary feeding mechanism includes a gear shaft rotatably arranged on the base, the gear shaft is driven to rotate by a motor, and drives the annular rotating material disk installed on the base to rotate, a plurality of material boxes are evenly arranged on the circumference of the rotating material disk, a gasket outlet is arranged on the outer side of the material box away from the center of the rotating material disk, a push plate push-in hole is arranged on the inner side of the material box close to the center of the rotating material disk, a push cylinder is installed on the base and located in the annular rotating material disk, a push plate matching with the push plate push-in hole is arranged on the end of the piston rod of the push cylinder, when the push cylinder drives the push plate to extend, the push plate enters the material box through the push plate push-in hole, and pushes the gasket in the material box into the two-way positioning feeding mechanism through the gasket outlet;

[0007] The bidirectional positioning feeding mechanism comprises a feeding module arranged on a base, a base plate is arranged on a mobile platform of the feeding module, a positioning base plate is arranged on the base plate, a tightening cylinder is arranged on the tail of the positioning base plate, a tightening block is arranged on the piston rod of the tightening cylinder, a self-locking rod parallel to the piston rod of the tightening cylinder is arranged on the lower part of the tightening block on the side away from the tightening cylinder, and an L-shaped rotating positioning block driven and rotated by the self-locking rod is hinged on the head of the positioning base plate.

[0008] Mounting seats are respectively provided on both sides of the positioning base plate, and sliding seats are provided on the mounting seats. Side positioning blocks are provided on the sliding seats on both sides, and the distance between the two side positioning blocks is equal to the width of the gasket. An optical axis is provided on the side positioning block, and a spring is sleeved on the optical axis. The tail of the optical axis movably passes through the optical axis support arranged on the base plate; a gasket slideway is provided on the side positioning block close to the rotating feeding mechanism, and the upper surface of the gasket slideway is flush with the upper surface of the positioning base plate; a limiting convex block matching with the end face of the mounting seat is provided below the end of the side positioning block close to the spring, and under the action of the spring, the limiting convex block of the side positioning block presses against the end face of the mounting seat so that the side positioning block keeps its position fixed;

[0009] The three-axis visual riveting mechanism includes an X-axis module arranged on a base, on which a Y-axis module is provided; on the Y-axis module, a Z-axis module is provided, on which a rivet gun mounting seat is provided, on which a rivet gun is fixed, and on which a CCD camera and a light source electrically connected to a control center are also provided.

[0010] As a preferred solution, a slot with an inverted T-shaped cross section is provided on the mounting seat, and a clamping portion matching the slot is provided at the lower portion of the sliding seat.

[0011] As a preferred solution, a self-locking rod guide block for the self-locking rod to pass through is provided at the bottom of the positioning base plate;

[0012] The technical problem to be solved by the present invention is to provide a feeding and riveting method for an elevator guide rail gasket with continuous automatic feeding and an intelligent positioning riveting device.

[0013] In order to solve the above technical problems, the method adopted by the present invention is: the feeding and riveting method of the elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device as described in any of the above items, comprising the following steps:

[0014] Step 1: Positioning the guide rail riveting holes

[0015] 1.1 Image acquisition

[0016] Control the movement of the three-axis visual riveting mechanism, drive the CCD camera to the elevator guide rail photography position, make the elevator guide rail riveting hole within the field of view of the CCD camera, and collect the image of the elevator guide rail end through the CCD camera;

[0017] 1.2 Template Matching

[0018] The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing;

[0019] 1.3 Image Preprocessing

[0020] The source image of step 1.2 is subjected to denoising by using a combined filtering method; the combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set, and after the combined filtering is performed, a denoised image is generated; the details are as follows:

[0021] a. Set the value range of each filter kernel size: the mean filter kernel size is Ka, the Gaussian filter kernel size is Kb, the median filter kernel size is Kc, and the bilateral filter kernel size is Kd, and the value range is [1, 20];

[0022] b. In one optimization, the kernel size of each filter is averaged from the value range to obtain m values, and each filter takes a kernel size value to obtain m 4 filter kernel size combination value, using m 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain m 4 After denoising, calculate m 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [Ka′, Kb′, Kc′, Kd′];

[0023] c. Secondary optimization, expand the range of the filter kernel size combination value [Ka′, Kb′, Kc′, Kd′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [Ka′-T, Ka′+T]; the Gaussian filter kernel value range is [Kb′-T, Kb′+T]; the median filter kernel value range is [Kc′-T, Kc′+T]; the bilateral filter kernel value range is [Kd′-T, Kd′+T], T is [0.2, 1.2]; take an average of n values ​​of each filter kernel value from the new value range to obtain n 4 filter kernel combination value, using n 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain n 4 After denoising, calculate n 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing;

[0024] 1.4 Image Enhancement

[0025] Binarization and morphological operations are performed on the source image preprocessed in step 1.3. The morphological operations include image erosion and dilation. The eroded image is dilated to remove the interference area and keep the original hole area contour unchanged.

[0026] 1.5 Hole position identification

[0027] The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system;

[0028] 1.7 Hole coordinate transformation

[0029] The coordinates of the rivet holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the three-axis visual riveting mechanism coordinate system. The coordinates of the two guide rail rivet holes are (x1, y1) and (x2, y2), respectively. The coordinate transformation matrix is ​​obtained during the CCD camera calibration process.

[0030] Step 2: Gasket filling

[0031] The push cylinder extends to drive the push plate to extend, and the push plate enters the material box through the push plate push hole, pushes the gasket in the material box through the gasket outlet, and enters the two-way positioning feeding mechanism; if the gaskets in the current position of the material box are pushed out, the motor drives the gear shaft to rotate, drives the rotating material tray to rotate, and aligns the push plate push hole of the next position of the material box with the push plate;

[0032] Step 3: Bidirectional positioning of gasket

[0033] When the gasket is pushed out from the rotary feeding mechanism, it passes through the gasket slideway and enters the bidirectional positioning feeding mechanism. At this time, the tightening cylinder extends out, driving the tightening block to move forward, and the tightening block pushes the gasket to move forward. During the movement of the gasket, the lateral position accuracy is guaranteed under the limiting conditions of the sliding positioning blocks on both sides; the tightening block drives the self-locking rod to move forward, and before the gasket reaches the position of the rotating positioning block, the self-locking rod lifts and rotates the rotating positioning block and locks it to a vertical state, and the tightening block pushes the gasket to tighten the rotating positioning block to achieve the longitudinal positioning of the gasket;

[0034] Step 4: Position the gasket rivet holes

[0035] 4.1 Image acquisition

[0036] After the bidirectional positioning of the gasket is completed, the feeding module drives the gasket to move to the photographing position, and moves the CCD camera backward along the Y direction by a distance ε to the photographing position;

[0037] 4.2 Template Matching

[0038] The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing;

[0039] 4.3 Image Preprocessing

[0040] The image in step 4.2 is subjected to denoising by using a combined filtering method; the combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set, and after the combined filtering is performed, a denoised image is generated, as follows:

[0041] a. Set the value range of each filter kernel size. The mean filter kernel size is La, the Gaussian filter kernel size is Lb, the median filter kernel size is Lc, and the bilateral filter kernel size is Ld. The value range is [1, 20];

[0042] b. In one optimization, the kernel size of each filter is averaged from the value range to obtain i values, and each filter takes a kernel size value to obtain i 4 filter kernel size combination value, using i 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain i 4 After denoising, calculate i 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [La′, Lb′, Lc′, Ld′];

[0043] c. Secondary optimization, expand the range of the filter kernel size combination value [La′, Lb′, Lc′, Ld′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [La′-P, La′+P]; the Gaussian filter kernel value range is [Lb′-P, Lb′+P]; the median filter kernel value range is [Lc′-P, Lc′+P]; the bilateral filter kernel value range is [Ld′-P, Ld′+P], P is [0.2, 1.2]; take j values ​​of each filter kernel value from the average value range, and get j 4 filter kernel combination value, using j 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters, and the image is denoised to obtain j 4 After denoising, calculate j 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing;

[0044] 4.4 Image Enhancement

[0045] Binarization and morphological operations are performed on the preprocessed image. Morphological operations include image erosion and dilation. Dilation operations are performed on the eroded image to remove interference areas and keep the original hole area contour unchanged.

[0046] 4.5 Hole Position Identification

[0047] The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system;

[0048] 4.6 Hole Coordinate Transformation

[0049] The coordinates of the rivet holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the three-axis visual riveting mechanism coordinate system. The coordinates of the two gasket rivet holes are (Px1, Py1) and (Px2, Py2) respectively.

[0050] Step 5: Feeding the gasket

[0051] Calculate the longitudinal difference δ between the position coordinates of the guide rail riveting hole and the gasket riveting hole as the feed amount of the feeding module to ensure that the guide rail riveting hole and the gasket riveting hole are accurately aligned;

[0052] Among them, the longitudinal difference

[0053] Step 6: Automatic riveting

[0054] The three-axis visual riveting mechanism drives the rivet gun to move the riveting hole position according to the coordinates (x1, y1) and (x2, y2) of the riveting hole of the guide rail to complete the riveting of the gasket;

[0055] Step 7: Reset the positioning mechanism

[0056] After riveting is completed, the tightening cylinder retracts, driving the self-locking rod to retract. At this time, the rotating positioning block can rotate, releasing the locked state of the rotating positioning block. At this time, the feeding module drives the positioning mechanism to retract, and the sliding positioning blocks on both sides are reset under the action of the springs.

[0057] As a preferred solution, m is any integer from 5 to 10, n is any integer from 5 to 10, i is any integer from 5 to 10, and j is any integer from 5 to 10.

[0058] The beneficial effects of the present invention are:

[0059] This device adopts a self-locking longitudinal positioning mechanism to ensure the longitudinal positioning accuracy of the gasket. Due to the small internal space of the guide rail, it is impossible to add a power device at the end of the longitudinal positioning mechanism to realize the movement of the longitudinal positioning block. Therefore, a self-locking structure is adopted. The longitudinal positioning and tightening of the gasket can be achieved by only the telescopic movement of the tightening cylinder to drive the longitudinal positioning block to rotate. No additional power device is needed. At the same time, after the riveting of the gasket is completed, the positioning self-locking is released, so that the positioning device can be smoothly withdrawn from the guide rail; due to the use of a telescopic transverse positioning mechanism, the transverse position of the gasket is not offset during the longitudinal positioning process of the gasket, thereby ensuring the alignment accuracy of the riveting holes of the gasket and the guide rail in the transverse degree of freedom; during the process of the gasket entering the guide rail, the transverse positioning device retracts to avoid the transverse positioning structure being unable to enter the narrow space inside the guide rail, resulting in a large transverse positioning error of the gasket.

[0060] This feeding riveting method adopts a visual processing method to locate the rivet hole positions of the guide rail and the gasket respectively, and calculates the difference between the rivet holes of the guide rail and the gasket in the longitudinal degree of freedom as the longitudinal feed amount of the feeding, thereby ensuring the alignment accuracy of the rivet holes of the gasket and the guide rail in the longitudinal degree of freedom; at the same time, the position coordinates of the rivet holes on the guide rail are used as the moving coordinates of the automatic riveting device, thereby ensuring the accuracy of the riveting position.

[0061] The feeding riveting method adopts a combined filtering method to process the collected workpiece image, and adopts a secondary optimization method to find the optimal filtering combination, thereby reducing the amount of image processing calculation, improving the image filtering effect, and facilitating the recognition accuracy of the workpiece hole position. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a diagram of a gasket positioning device of the present invention;

[0063] Figure 2 This is a schematic diagram of positioning and riveting of the gasket of the present invention;

[0064] Figure 3 This is a diagram of the rotary feeding mechanism of the present invention;

[0065] Figure 4This is a diagram of a bidirectional positioning feeding mechanism of the present invention;

[0066] Figure 5 It is a partial view of the bidirectional positioning feeding mechanism of the present invention;

[0067] Figure 6 This is a self-locking state diagram of the longitudinal positioning mechanism of the present invention;

[0068] Figure 7 This is a diagram of the three-axis visual riveting mechanism of the present invention.

[0069] Figure 8 This is a flow chart of the riveting hole identification and positioning of the present invention

[0070] Figures 1 to 8 middle:

[0071] 1 base; 2 rotating feeding mechanism; 21 gear shaft; 22 rotating material tray; 23 material box; 231 gasket discharge port; 232 push plate push hole; 24 push cylinder; 25 push plate;

[0072] 3 bidirectional positioning feeding mechanism; 31 feeding module; 32 bottom plate; 33 positioning bottom plate; 34 tightening cylinder; 35 tightening block; 36 rotating positioning block; 37 side sliding positioning block; 38 side sliding positioning block; 39 gasket slideway; 310 optical axis; 311 optical axis support; 312 self-locking rod; 313 self-locking rod guide block; 314 slide seat; 315 mounting seat;

[0073] 4 three-axis visual riveting mechanism; 41 X-axis module; 42 Y-axis module; 43 Z-axis module; 44 rivet gun mounting seat; 45 rivet gun; 46 CCD camera; 47 light source; 5 elevator guide rail; 51 guide rail riveting hole; 6 gasket; 61 gasket riveting hole. DETAILED DESCRIPTION

[0074] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0075] like Figure 1 As shown, the elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device includes a base 1, and a rotating feeding mechanism 2, a bidirectional positioning feeding mechanism 3 and a three-axis visual riveting mechanism 4 arranged on the base;

[0076] like Figure 2 As shown, the device is used to accurately deliver the gasket 6 to the inside of the elevator guide rail 5 and align the guide rail riveting hole 51 and the gasket riveting hole 61;

[0077] like Figure 3As shown, the rotary feeding mechanism 2 includes a gear shaft 21 rotatably arranged on the base, the gear shaft 21 is driven to rotate by a motor, and drives the annular rotating material disc 22 installed on the base 1 to rotate, and a plurality of material boxes 23 are evenly arranged on the circumferential direction of the rotating material disc 22, and a gasket discharge port 231 is arranged on the outer side of the material box 23 away from the center of the rotating material disc, and a push plate push-in hole 232 is arranged on the inner side of the material box 23 close to the center of the rotating material disc, and a push cylinder 24 is installed on the base and located in the annular rotating material disc, and a push plate 25 matching with the push plate push-in hole 232 is arranged on the piston rod end of the push cylinder 24, and when the push cylinder 24 drives the push plate 25 to extend, the push plate 25 enters the material box 23 through the push plate push-in hole 232, and pushes the gasket 6 in the material box 23 into the two-way positioning feeding mechanism 3 through the gasket discharge port 231;

[0078] like Figures 4 to 6 As shown, the bidirectional positioning feeding mechanism 3 includes a feeding module 31 arranged on the base 1, a base plate 32 is arranged on the mobile platform of the feeding module 31, a positioning base plate 33 is arranged on the base plate 32, a tightening cylinder 34 is arranged on the tail of the positioning base plate 33, a tightening block 35 is arranged on the piston rod of the tightening cylinder 34, a self-locking rod 312 parallel to the piston rod of the tightening cylinder 34 is arranged on the side of the lower part of the tightening block 35 away from the tightening cylinder 34, an L-shaped rotating positioning block 36 driven and rotated by the self-locking rod 312 is hinged on the head of the positioning base plate 33, and a self-locking rod guide block 313 for the self-locking rod 312 to pass through is arranged at the bottom of the positioning base plate 33;

[0079] Mounting seats 315 are respectively provided on both sides of the positioning base plate 33 , and a slide seat 314 is provided on the mounting seat 315 . A clamping groove with an inverted T-shaped cross section is provided on the mounting seat 315 , and a clamping portion matching the clamping groove is provided at the lower part of the slide seat 314 . Side positioning blocks 37 and 38 are provided on the slide seats 314 on both sides, and the distance between the two side positioning blocks 37 and 38 is equal to the width of the gasket. The side positioning blocks 37 and 38 are provided with an optical axis 310, and a spring 316 is sleeved on the optical axis 310. The tail of the optical axis 310 movably passes through the optical axis support 311 set on the base plate 32; a gasket slideway 39 is provided on the side positioning block 38 close to the rotary feeding mechanism 2, and the upper surface of the gasket slideway 39 is flush with the upper surface of the positioning bottom plate 33; a limiting protrusion matching the end face of the mounting seat 315 is provided below the end of the side positioning blocks 37 and 38 close to the spring 316, and the limiting protrusion of the side positioning blocks 37 and 38 is pressed against the end face of the mounting seat 315 under the action of the spring 316 so that the side positioning blocks 37 and 38 keep their positions fixed;

[0080] The three-axis visual riveting mechanism 4 includes an X-axis module 41 arranged on the base 1, and a Y-axis module 42 is provided on the X-axis module 41; a Z-axis module 43 is provided on the Y-axis module 42, and a rivet gun mounting seat 44 is provided on the Z-axis module 43, and a rivet gun 45 is fixed on the rivet gun mounting seat, and a CCD camera 46 and a light source 47 electrically connected to the control center are also provided on the Z-axis module 43.

[0081] When the gasket 6 is pushed out from the rotary feeding mechanism 2, the gasket 6 passes through the gasket slide 39 and enters the bidirectional positioning feeding mechanism 3. At this time, the tightening cylinder 34 extends out, driving the tightening block 35 to move forward, and the tightening block 35 pushes the gasket 6 to move forward. During the movement of the gasket 6, the lateral position accuracy is guaranteed under the limit of the two side positioning blocks 37 and 38; at the same time, the tightening block 35 drives the self-locking rod 312 to move forward. Before the gasket reaches the position of the rotating positioning block 36, the self-locking rod 312 lifts and rotates the rotating positioning block 36 until it is in a vertical state, as shown in FIG. Figure 6 As shown, at this time, the tightening block 35 pushes the gasket 6 to tighten the rotating positioning block 36 to achieve the longitudinal positioning of the gasket 6;

[0082] After the gasket 6 is positioned and locked, the feeding module 31 drives the base plate 32 to move forward and delivers the gasket 6 to the inside of the hollow elevator guide rail 5. When the gasket 6 enters the inside of the elevator guide rail 5, the two side sliding positioning blocks 37 and 38 contact the outer contour of the elevator guide rail 5, and the outer contour of the elevator guide rail 5 compresses the positioning side sliding positioning blocks 37 and 38 to move backward; when the riveting is completed, the tightening cylinder 34 retracts, driving the self-locking rod 312 to retract, releasing the locked state of the rotating positioning block 36, and the rotating positioning block 36 rotates downward, and the feeding module 31 drives the positioning mechanism to retract, and the positioning blocks 37 and 38 on both sides are reset under the action of the spring;

[0083] like Figure 7 As shown, the three-axis visual riveting mechanism 4 includes an X-axis module 41, on which a Y-axis module 42 is mounted; on the Y-axis module 42, a Z-axis module 43 is mounted, on which a rivet gun mounting seat 44 is mounted, on which a rivet gun 45 is fixed, and on which a CCD camera 46 and a light source 47 are also mounted.

[0084] like Figure 8 As shown, the feeding and riveting method of the elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device comprises the following steps:

[0085] Step 1: Positioning the guide rail riveting holes

[0086] 1.1 Image acquisition

[0087] Control the movement of the three-axis visual riveting mechanism 4 to drive the CCD camera 46 to the elevator guide rail photographing position, so that the elevator guide rail riveting hole 51 is within the field of view of the CCD camera 46, and the end image of the elevator guide rail 5 is collected by the CCD camera 46;

[0088] 1.2 Template Matching

[0089] The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing;

[0090] 1.3 Image Preprocessing

[0091] The source image of step 1.2 is subjected to denoising by using a combined filtering method; the combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set, and after the combined filtering is performed, a denoised image is generated; the details are as follows:

[0092] a. Set the value range of each filter kernel size: the mean filter kernel size is Ka, the Gaussian filter kernel size is Kb, the median filter kernel size is Kc, and the bilateral filter kernel size is Kd, and the value range is [1, 20];

[0093] b. In one optimization, the kernel size of each filter is averaged from the value range to obtain m values, where m is any integer between 5 and 10. Each filter takes a kernel size value and obtains m. 4 filter kernel size combination value, using m 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain m 4 After denoising, calculate m 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [Ka′, Kb′, Kc′, Kd′];

[0094] c. Secondary optimization, expand the range of the filter kernel size combination value [Ka′, Kb′, Kc′, Kd′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [Ka′-T, Ka′+T]; the Gaussian filter kernel value range is [Kb′-T, Kb′+T]; the median filter kernel value range is [Kc′-T, Kc′+T]; the bilateral filter kernel value range is [Kd′-T, Kd′+T], T is [0.2, 1.2]; take an average of n values ​​of each filter kernel value from the new value range, n is any integer from 5 to 10, and get n 4 filter kernel combination value, using n 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain n 4 After denoising, calculate n 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing;

[0095] 1.4 Image Enhancement

[0096] Binarization and morphological operations are performed on the source image preprocessed in step 1.3. The morphological operations include image erosion and dilation. The eroded image is dilated to remove the interference area and keep the original hole area contour unchanged.

[0097] 1.5 Hole position identification

[0098] The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system;

[0099] 1.7 Hole coordinate transformation

[0100] The coordinates of the riveting holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the three-axis visual riveting mechanism 4 coordinate system. The position coordinates of the two guide rail riveting holes 51 are (x1, y1) and (x2, y2), respectively. The coordinate transformation matrix is ​​obtained during the calibration process of the CCD camera 46.

[0101] Step 2: Gasket filling

[0102] The push cylinder 24 extends to drive the push plate 25 to extend, and the push plate 25 enters the material box 23 through the push plate push hole 232, pushes the gasket 6 in the material box 23 through the gasket outlet 231, and enters the two-way positioning feeding mechanism 3; if the gasket 6 in the material box 23 at the current position is pushed out, the motor drives the gear shaft 21 to rotate, drives the rotating material tray 22 to rotate, and aligns the push plate push hole 232 of the material box 23 at the next position with the push plate 25;

[0103] Step 3: Bidirectional positioning of gasket

[0104] When the gasket 6 is pushed out from the rotary feeding mechanism 2, the gasket 6 enters the bidirectional positioning feeding mechanism 3 through the gasket slide 39. At this time, the tightening cylinder 34 extends out, driving the tightening block 35 to move forward, and the tightening block 35 pushes the gasket 6 to move forward. During the movement of the gasket 6, the lateral position accuracy is guaranteed under the limiting conditions of the sliding positioning blocks on both sides; at the same time, the tightening block 35 drives the self-locking rod 312 to move forward. Before the gasket reaches the position of the rotating positioning block 36, the self-locking rod 312 lifts and rotates the rotating positioning block 36 and locks it to a vertical state. Figure 7 As shown, at this time, the pressing block 35 can continue to move forward, pushing the gasket 6 to press against the rotating positioning block 36, so as to realize the longitudinal positioning of the gasket 6;

[0105] Step 4: Position the gasket rivet holes

[0106] 4.1 Image acquisition

[0107] After the two-way positioning of the gasket 6 is completed, the feeding module 31 drives the gasket 6 to move to the photographing position. The front end surface of the gasket 6 at the photographing position is 5 to 15 mm away from the end surface of the elevator guide rail 5. In order to make the position of the gasket riveting hole 61 within the field of view of the CCD camera 46, the CCD camera 46 must be moved backward along the Y direction by a distance ε, and ε is set according to the on-site conditions;

[0108] 4.2 Template Matching

[0109] The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing;

[0110] 4.3 Image Preprocessing

[0111] Use combined filtering to reduce noise on the image in step 4.2;

[0112] The combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set. After the combined filtering is performed, a denoised image is generated. The specific method is as follows:

[0113] a. Set the value range of each filter kernel size, as follows;

[0114] The kernel size of mean filter La, Gaussian filter Lb, median filter Lc, and bilateral filter Ld all range from [1, 20].

[0115] b. In one optimization, the kernel size of each filter is averaged from the value range to obtain i values, i is 5 to 10, and each filter takes a kernel size value to obtain i 4 filter kernel size combination value, using i 4 The combined values ​​of the filter kernel sizes are used as the combined filter parameters to perform denoising on the image, and m4 denoised images are obtained. 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [La′, Lb′, Lc′, Ld′];

[0116] c. Secondary optimization, expand the range of the filter kernel size combination value [La′, Lb′, Lc′, Ld′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [La′-P, La′+P]; the Gaussian filter kernel value range is [Lb′-P, Lb′+P]; the median filter kernel value range is [Lc′-P, Lc′+P]; the bilateral filter kernel value range is [Ld′-P, Ld′+P], P is [0.2, 1.2]; take j values ​​of each filter kernel value from the value range of the center, j is 5 to 10, and get j4 filter kernel combination value, using j 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters, and the image is denoised to obtain j 4 After denoising, calculate j 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing;

[0117] 4.4 Image Enhancement

[0118] Binarization and morphological operations are performed on the preprocessed image. Morphological operations include image erosion and dilation. Dilation operations are performed on the eroded image to remove interference areas and keep the original hole area contour unchanged.

[0119] 4.5 Hole Position Identification

[0120] The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system;

[0121] 4.6 Hole Coordinate Transformation

[0122] The coordinates of the rivet holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the coordinate system of the three-axis visual riveting mechanism 4. The position coordinates of the two gasket rivet holes 61 are (Px1, Py1) and (Px2, Py2) respectively.

[0123] Step 5: Feeding the gasket

[0124] Calculate the longitudinal difference δ between the position coordinates of the guide rail riveting hole 51 and the gasket riveting hole 61 as the feed amount of the feeding module 31 to ensure that the guide rail riveting hole 51 and the gasket riveting hole 61 are accurately aligned;

[0125] Among them, the longitudinal difference

[0126] Step 6: Automatic riveting

[0127] The three-axis visual riveting mechanism 4 drives the rivet gun 45 to move the riveting hole position according to the coordinates (x1, y1) and (x2, y2) of the guide rail riveting hole 51 to complete the gasket riveting;

[0128] Step 7: Reset the positioning mechanism

[0129] After riveting is completed, the tightening cylinder 34 retracts, driving the self-locking rod 312 to retract. At this time, the rotating positioning block 36 can rotate, and the locked state of the rotating positioning block 36 is released. At this time, the feeding module 31 drives the positioning mechanism to retreat, and the sliding positioning blocks 37 and 38 on both sides are reset under the action of the springs.

[0130] The above-mentioned embodiments are only illustrative of the principles and effects of the invention, as well as some embodiments of its application, and are not intended to limit the invention. It should be pointed out that a person skilled in the art can make several modifications and improvements without departing from the inventive concept of the invention, and all of these belong to the protection scope of the invention.

Claims

1. Elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device, characterized by: It includes a base, and a rotating feeding mechanism, a bidirectional positioning feeding mechanism and a three-axis visual riveting mechanism arranged on the base; The rotary feeding mechanism includes a gear shaft rotatably arranged on the base, the gear shaft is driven to rotate by a motor, and drives the annular rotating material disk installed on the base to rotate, a plurality of material boxes are evenly arranged on the circumference of the rotating material disk, a gasket outlet is arranged on the outer side of the material box away from the center of the rotating material disk, a push plate push-in hole is arranged on the inner side of the material box close to the center of the rotating material disk, a push cylinder is installed on the base and located in the annular rotating material disk, a push plate matching with the push plate push-in hole is arranged on the end of the piston rod of the push cylinder, when the push cylinder drives the push plate to extend, the push plate enters the material box through the push plate push-in hole, and pushes the gasket in the material box into the two-way positioning feeding mechanism through the gasket outlet; The bidirectional positioning feeding mechanism comprises a feeding module arranged on a base, a base plate is arranged on a mobile platform of the feeding module, a positioning base plate is arranged on the base plate, a tightening cylinder is arranged on the tail of the positioning base plate, a tightening block is arranged on the piston rod of the tightening cylinder, a self-locking rod parallel to the piston rod of the tightening cylinder is arranged on the lower part of the tightening block on the side away from the tightening cylinder, and an L-shaped rotating positioning block driven and rotated by the self-locking rod is hinged on the head of the positioning base plate. Mounting seats are respectively provided on both sides of the positioning base plate, and sliding seats are provided on the mounting seats. Side positioning blocks are provided on the sliding seats on both sides, and the distance between the two side positioning blocks is equal to the width of the gasket. An optical axis is provided on the side positioning block, and a spring is sleeved on the optical axis. The tail of the optical axis movably passes through the optical axis support arranged on the base plate; a gasket slideway is provided on the side positioning block close to the rotating feeding mechanism, and the upper surface of the gasket slideway is flush with the upper surface of the positioning base plate; a limiting convex block matching with the end face of the mounting seat is provided below the end of the side positioning block close to the spring, and under the action of the spring, the limiting convex block of the side positioning block presses against the end face of the mounting seat so that the side positioning block keeps its position fixed; The three-axis visual riveting mechanism includes an X-axis module arranged on a base, on which a Y-axis module is provided; on the Y-axis module, a Z-axis module is provided, on which a rivet gun mounting seat is provided, on which a rivet gun is fixed, and on which a CCD camera and a light source electrically connected to a control center are also provided.

2. The elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device according to claim 1, characterized in that: The mounting seat is provided with a clamping groove with an inverted T-shaped cross section, and the lower part of the sliding seat is provided with a clamping portion matched with the clamping groove.

3. The elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device according to claim 2, characterized in that: A self-locking rod guide block for the self-locking rod to pass through is provided at the bottom of the positioning base plate.

4. A feeding and riveting method for the elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device according to any one of claims 1 to 3, comprising the following steps: Step 1: Positioning the guide rail riveting holes 1.1 Image acquisition Control the movement of the three-axis visual riveting mechanism, drive the CCD camera to the elevator guide rail photography position, make the elevator guide rail riveting hole within the field of view of the CCD camera, and collect the image of the elevator guide rail end through the CCD camera; 1.2 Template Matching The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing; 1.3 Image Preprocessing The source image of step 1.2 is subjected to denoising by using a combined filtering method; the combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set, and after the combined filtering is performed, a denoised image is generated; the details are as follows: a. Set the value range of each filter kernel size: the mean filter kernel size is Ka, the Gaussian filter kernel size is Kb, the median filter kernel size is Kc, and the bilateral filter kernel size is Kd, and the value range is [1, 20]; b. In one optimization, the kernel size of each filter is averaged from the value range to obtain m values, and each filter takes a kernel size value to obtain m 4 filter kernel size combination value, using m 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain m 4 After denoising, calculate m 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [Ka′, Kb′, Kc′, Kd′]; c. Secondary optimization, expand the range of the filter kernel size combination value [Ka′, Kb′, Kc′, Kd′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [Ka′-T, Ka′+T]; the Gaussian filter kernel value range is [Kb′-T, Kb′+T]; the median filter kernel value range is [Kc′-T, Kc′+T]; the bilateral filter kernel value range is [Kd′-T, Kd′+T], T is [0.2, 1.2]; take an average of n values ​​of each filter kernel value from the new value range to obtain n 4 filter kernel combination value, using n 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain n 4 After denoising, calculate n 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing; 1.4 Image Enhancement Binarization and morphological operations are performed on the source image preprocessed in step 1.

3. The morphological operations include image erosion and dilation. The eroded image is dilated to remove the interference area and keep the original hole area contour unchanged. 1.5 Hole position identification The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system; 1.7 Hole coordinate transformation The coordinates of the rivet holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the three-axis visual riveting mechanism coordinate system. The coordinates of the two guide rail rivet holes are (x1, y1) and (x2, y2), respectively. The coordinate transformation matrix is ​​obtained during the CCD camera calibration process. Step 2: Gasket filling The push cylinder extends to drive the push plate to extend, and the push plate enters the material box through the push plate push hole, pushes the gasket in the material box through the gasket outlet, and enters the two-way positioning feeding mechanism; if the gaskets in the current position of the material box are pushed out, the motor drives the gear shaft to rotate, drives the rotating material tray to rotate, and aligns the push plate push hole of the next position of the material box with the push plate; Step 3: Bidirectional positioning of gasket When the gasket is pushed out from the rotary feeding mechanism, it passes through the gasket slideway and enters the bidirectional positioning feeding mechanism. At this time, the tightening cylinder extends out, driving the tightening block to move forward, and the tightening block pushes the gasket to move forward. During the movement of the gasket, the lateral position accuracy is guaranteed under the limiting conditions of the sliding positioning blocks on both sides; the tightening block drives the self-locking rod to move forward, and before the gasket reaches the position of the rotating positioning block, the self-locking rod lifts and rotates the rotating positioning block and locks it to a vertical state, and the tightening block pushes the gasket to tighten the rotating positioning block to achieve the longitudinal positioning of the gasket; Step 4: Position the gasket rivet holes 4.1 Image acquisition After the bidirectional positioning of the gasket is completed, the feeding module drives the gasket to move to the photographing position, and moves the CCD camera backward along the Y direction by a distance ε to the photographing position; 4.2 Template Matching The collected image is grayed out, the target area is determined by template matching, and the target area is cut out as the source image for subsequent hole position recognition processing; 4.3 Image Preprocessing The image in step 4.2 is subjected to denoising by using a combined filtering method; the combined filtering includes mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The kernel size of each filter is set, and after the combined filtering is performed, a denoised image is generated, as follows: a. Set the value range of each filter kernel size. The mean filter kernel size is La, the Gaussian filter kernel size is Lb, the median filter kernel size is Lc, and the bilateral filter kernel size is Ld. The value range is [1, 20]; b. In one optimization, the kernel size of each filter is averaged from the value range to obtain i values, and each filter takes a kernel size value to obtain i 4 filter kernel size combination value, using i 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters to perform denoising on the image and obtain i 4 After denoising, calculate i 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected. The corresponding filter kernel size combination value is recorded as [La′, Lb′, Lc′, Ld′]; c. Secondary optimization, expand the range of the filter kernel size combination value [La′, Lb′, Lc′, Ld′] obtained in the previous step to obtain a new filter kernel size range: the mean filter kernel value range is [La′-P, La′+P]; the Gaussian filter kernel value range is [Lb′-P, Lb′+P]; the median filter kernel value range is [Lc′-P, Lc′+P]; the bilateral filter kernel value range is [Ld′-P, Ld′+P], P is [0.2, 1.2]; take j values ​​of each filter kernel value from the average value range, and get j 4 filter kernel combination value, using j 4 The combined values ​​of the filter kernel sizes are used as combined filter parameters, and the image is denoised to obtain j 4 After denoising, calculate j 4 The signal-to-noise ratio of the denoised images is calculated, and the image with the largest signal-to-noise ratio is selected as the source image for subsequent processing; 4.4 Image Enhancement Binarization and morphological operations are performed on the preprocessed image. Morphological operations include image erosion and dilation. Dilation operations are performed on the eroded image to remove interference areas and keep the original hole area contour unchanged. 4.5 Hole Position Identification The contour recognition method is used to identify the hole features in the image and obtain the coordinates of the riveting holes in the image coordinate system; 4.6 Hole Coordinate Transformation The coordinates of the rivet holes in the image coordinate system are multiplied by the coordinate transformation matrix to obtain the coordinates in the three-axis visual riveting mechanism coordinate system. The coordinates of the two gasket rivet holes are (Px1, Py1) and (Px2, Py2) respectively. Step 5: Feeding the gasket Calculate the longitudinal difference δ between the position coordinates of the guide rail riveting hole and the gasket riveting hole as the feed amount of the feeding module to ensure that the guide rail riveting hole and the gasket riveting hole are accurately aligned; in, Longitudinal difference Step 6: Automatic riveting The three-axis visual riveting mechanism drives the rivet gun to move the riveting hole position according to the coordinates (x1, y1) and (x2, y2) of the riveting hole of the guide rail to complete the riveting of the gasket; Step 7: Reset the positioning mechanism After riveting is completed, the tightening cylinder retracts, driving the self-locking rod to retract. At this time, the rotating positioning block can rotate, releasing the locked state of the rotating positioning block. At this time, the feeding module drives the positioning mechanism to retract, and the sliding positioning blocks on both sides are reset under the action of the springs.

5. The feeding and riveting method of the elevator guide rail gasket continuous automatic feeding and intelligent positioning riveting device according to claim 4, characterized in that: The m is any integer from 5 to 10, the n is any integer from 5 to 10, i is any integer from 5 to 10, and j is any integer from 5 to 10.

Images