A round spring height batch detection device based on visual laser scanning

Abstract

The application discloses a kind of round spring height batch detection devices based on visual laser scanning, belong to spring detection technical field, including bearing table for placing the round spring to be detected;N standard rods are evenly distributed on the round spring bearing table;Motion mechanism is arranged above the spring bearing table, for realizing the translational motion mechanical device of laser scanning to the round spring to be measured;Motion control system is used to control the position and speed of scanning translational motion of motion mechanism;Probe is used to obtain single detection three-dimensional data of the round spring to be detected;Spring height calculation module is used to form spring three-dimensional point cloud data by combining motion information with single detection three-dimensional data collected by probe;Three-dimensional point cloud information is processed, and spring height is calculated.The device avoids the limitations of traditional contact measurement, improves spring batch detection efficiency and detection accuracy.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of spring detection, and relates to a round spring height batch detection device based on visual laser scanning. BACKGROUND

[0002] As a key elastic element in mechanical equipment, the height parameter of the round spring directly affects the assembly precision, bearing capacity and operation safety of the equipment. In core scenes such as railway freight car bogies, automobile suspensions and aerospace damping systems, the consistency and precision of the spring height are the core indicators to ensure the system performance.

[0003] The traditional round spring height detection method has significant technical bottlenecks:

[0004] Limitations of contact measurement: manual measurement relying on vernier calipers, height gauges and other tools, single spring detection takes 30-60 seconds, and the labor intensity is high. Detection personnel fatigue can lead to errors of more than ±0.5mm; mechanical contact gauges further reduce precision due to wear after long-term use, and need to be calibrated and maintained multiple times a year, increasing costs.

[0005] Low batch detection efficiency: in the traditional assembly line, each workstation can detect at most 200-300 springs per hour, which is difficult to meet the large-scale needs of railway freight car maintenance and automobile production lines, becoming a bottleneck of production capacity.

[0006] Data management defects: detection results rely on manual recording, which is prone to errors and omissions, and cannot be associated with production batches and maintenance cycles. When equipment fails, it is difficult to trace the source of spring quality problems, which restricts process optimization.

[0007] Complex spring detection problems: for combined springs with inner and outer nesting, traditional methods need to be split for measurement, which destroys the assembly state, and cannot synchronously obtain the height difference between the inner and outer springs, affecting the judgment of assembly adaptability.

[0008] The traditional method not only takes a long time, and manual measurement is prone to errors due to fatigue. Mechanical contact may also cause spring deformation due to contact force. The emergence of laser scanning visual measurement technology provides a possibility to solve the above problems. SUMMARY

[0009] To solve the above problems, a round spring height batch detection device based on visual laser scanning includes:

[0010] Spring bearing table: including a bearing table and N standard rods; the bearing table is used for placing the round spring to be detected; N standard rods: are uniformly distributed on the spring bearing table, and are used for fitting a reference plane for measuring the height of the round spring;

[0011] Motion module: including a motion mechanism and a motion control system;

[0012] Motion mechanism: arranged above the spring bearing table, for realizing the translational motion mechanical device of the laser scanning of the round spring to be tested;

[0013] Motion control system: for controlling the position and speed of the scanning translational motion of the motion mechanism;

[0014] Detection module: for obtaining a single detection three-dimensional data of the round spring to be detected;

[0015] Spring height calculation module: for calculating the three-dimensional coordinate data corresponding to the laser scanning line based on the single frame image transmitted by the probe based on the triangulation principle; and combining the motion information to form the spring three-dimensional point cloud data; processing the three-dimensional point cloud information to calculate the spring height.

[0016] Further: N standard rods: the preferred number of N is 9, and the 9 standard rods are evenly distributed on the upper surface of the bearing table in a 3x3 matrix, and the height of the standard rod is slightly greater than the height of the round spring to be detected.

[0017] Further: the detection module includes a probe, and the probe includes a first industrial camera, a second industrial camera and a line laser;

[0018] Line laser: for projecting onto the upper surface of the round spring to form a laser line;

[0019] First industrial camera and second industrial camera: for collecting the laser line image of the upper surface of the round spring in the entire detection process;

[0020] The first industrial camera and the second industrial camera are respectively arranged on both sides of the same horizontal line of the line laser.

[0021] Further: the motion mechanism includes:

[0022] Infrastructure: for serving as a support body;

[0023] Translation mechanism: arranged above the infrastructure, for moving the probe;

[0024] Drag chain mechanism: for accommodating various power lines and signal lines, and one end of the drag chain mechanism is fixed to the probe sliding table of the translation mechanism, and the drag chain mechanism guides the cable to move synchronously with the sliding table during the scanning process.

[0025] Further: the motion control system includes:

[0026] Servo motor: for driving the movement of the motion mechanism;

[0027] Servo driver: for driving the servo motor;

[0028] Motion controller: for controlling the servo driver;

[0029] First limit switch: provided at one end of the translation mechanism, for measuring the limit;

[0030] Second limit switch: provided at the other end of the translation mechanism, for measuring the limit.

[0031] Further, the detection module further comprises:

[0032] Panoramic camera: provided above the measured spring bearing table, for obtaining the overall image of the measured spring;

[0033] Illumination lamp: for providing light source for the panoramic camera shooting.

[0034] Further, based on the principle of triangulation, the three-dimensional coordinate data corresponding to the laser scanning line is calculated; and combined with the motion information, the spring three-dimensional point cloud data is formed; the spring height is calculated by processing the three-dimensional point cloud information as follows:

[0035] The internal parameters and external parameters of the first industrial camera and the second industrial camera are calibrated;

[0036] According to the external parameters of the calibrated first industrial camera and the second industrial camera, the conversion relationship between the camera coordinate system and the calibration coordinate system is obtained;

[0037] The plane equation of the laser plane in the calibration coordinate system is calibrated;

[0038] Based on the single-frame laser line image shot by the first industrial camera and the second industrial camera, the pixel points of the laser line are extracted from the image, which are converted to the camera coordinate system by using the internal parameter matrix, and then converted to the calibration coordinate system by using the coordinate conversion, and form a triangulation with the laser plane, calculate the three-dimensional point line corresponding to the single laser line of the first industrial camera and the second industrial camera, and fuse the three-dimensional point lines corresponding to the two cameras;

[0039] Convert the three-dimensional point line to the measurement coordinate system, and combine the movement vector to form the three-dimensional point cloud data of the spring and the standard rod;

[0040] The top surface data of the spring and the standard rod is extracted by segmenting the spring three-dimensional point cloud data;

[0041] The reference plane is fitted by using the top surface data of the standard rod, the distance from the top surface data of the spring to the reference plane is calculated, and the high points are selected according to the distance value in proportion, and the spring height is calculated by using the spring height algorithm.

[0042] Further: the process of calculating the spring height by using the spring height algorithm is as follows:

[0043] According to the obtained point cloud data, The top surface data of each circular spring center and the standard bar is extracted from the face projection data respectively;

[0044] A reference surface for measuring the height of the spring is obtained by plane fitting using the top surface data of the standard bar;

[0045] A ring-shaped area is screened according to the spring center point and the top surface of the spring to detect the to-be-detected point;

[0046] The distance from the to-be-detected point to the reference surface fitted by the top surface of the standard part is calculated;

[0047] The to-be-detected points are sorted in descending order according to the distance to the reference surface, and the points at the top are taken in proportion;

[0048] The points at the top are further plane fitted, and the ring range of the plane is determined;

[0049] The distance from the point on the plane in the ring to the nearest standard part is calculated;

[0050] The height of the spring is obtained by subtracting the nearest distance from the height of the standard bar.

[0051] Further, when the circular spring is a nested spring, the top surface data of the inner and outer springs is further divided, the inner and outer springs are divided by two ring-shaped areas respectively based on the spring center, and the heights of the inner and outer springs are calculated based on the spring height algorithm.

[0052] The device for batch detection of the height of a circular spring based on visual laser scanning, which adopts visual laser scanning technology to complete the batch detection task of the height of a circular spring, is based on non-contact characteristics, avoids the limitations of traditional contact measurement, and improves the batch detection efficiency and detection accuracy of the spring. At the same time, the detection data is uploaded to the management system in real time through the industrial network to form a full life cycle file containing the spring number, detection time and height parameter, providing data support for quality traceability and process improvement, and being particularly suitable for scenes with high safety and efficiency requirements in the railway, automobile and aerospace industries.

[0053] The device of the present application has the following effects:

[0054] The detection efficiency is significantly improved;

[0055] The spring detection scheme fully meets the detection needs of various industries with high efficiency and accuracy. In terms of detection efficiency, the detection of a single spring only takes 3-5 seconds, and the batch detection efficiency is 10-30 times higher than that of traditional manual or mechanical contact measurement. The scheme can quickly complete detection, effectively break through the production capacity bottleneck of traditional methods, and perfectly adapt to the efficient detection needs of large-scale production and maintenance in the railway, automobile and aerospace industries.

[0056] Technically, the left and right camera triangulation technology is adopted, and the imaging blind spot problem is completely solved through the fusion of double-view information. Single camera is prone to blind area due to high spring shielding low spring, depth of field limitation, etc. due to fixed view angle, resulting in data loss. The double cameras have a horizontal interval, which can be shot from different angles. The blind area of a certain view is often covered by another view, such as the inner spring area blocked by the outer spring of the left camera, which can be clearly captured by the right camera. Fusion of effective information from two perspectives can completely restore the inner and outer spring structure, avoid missing data caused by height difference, and ensure the accuracy and integrity of the height difference data.

[0057] In addition, the scheme supports batch detection of inner and outer spring sets, and related data can be obtained synchronously without disassembly. Traditional disassembly measurement not only has a complicated process and consumes a lot of time, but also may damage the spring during disassembly and reassembly, affecting its performance. The scheme eliminates the disassembly step, further improves the detection efficiency, and at the same time guarantees the integrity of the spring.

[0058] Improved detection accuracy and stability;

[0059] The detection accuracy is ±0.2mm, which is much higher than the accuracy level of traditional manual measurement, and can more accurately reflect the height parameters of the circular spring, ensuring the performance and assembly adaptability of the spring.

[0060] In the measurement mechanical structure, under the conditions of meeting the spring bearing platform flatness requirement of 0.05 and the length accuracy requirement of 0.02 of 9 standard rods, the standard rod is used to fit the reference surface, the translation mechanism realizes high-precision translation through servo motor and synchronous belt, and the control system realizes the collaborative work of each component, reducing the stability problem caused by tool wear, manual operation error, etc. in traditional contact measurement, and ensuring the stability and reliability of the detection results.

[0061] Realize full automation and intelligentization;

[0062] From the measurement device reset, spring placement, parameter setting to start measurement, result calculation and display, the whole process realizes automatic operation, reduces manual intervention, reduces labor intensity, and avoids errors caused by manual operation.

[0063] The detection software interface can automatically segment, number and associate the spring with the measurement data. The data display area clearly presents each parameter, and the operation area can conveniently set the measurement parameters, improving the intelligentization and convenience of operation.

[0064] Data traceability and quality control are strengthened;

[0065] The detection data can be uploaded to the management platform in real time to form a full life cycle database, including spring measurement results, detection time, etc. It is convenient for subsequent tracing of spring quality and provides a basis for quality problem investigation.

[0066] Through the statistics, query and analysis of historical data, data support can be provided for production process optimization, helping enterprises to improve product quality, especially suitable for industries with high safety and reliability requirements, to ensure the performance stability of key parts of the system.

[0067] Adapt to complex detection scenarios

[0068] Compared with traditional contact measurement, laser scanning visual measurement is a non-contact mode, which will not cause damage to the spring surface, and can detect complex structure springs such as inner and outer nesting, thereby expanding the detection application range.

[0069] The design of the translation mechanism can realize linear laser scanning translation, expand the visual detection range, and further improve the applicability and efficiency of detection. BRIEF DESCRIPTION OF DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiment or prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0071] Figure 1 It is the first angle structure schematic diagram of the round spring height batch detection device based on visual laser scanning;

[0072] Figure 2 It is the second angle structure schematic diagram of the round spring height batch detection device based on visual laser scanning;

[0073] Figure 3 It is the motion control system schematic diagram of the round spring height batch detection device based on visual laser scanning;

[0074] Figure 4 It is the software module working interface;

[0075] Figure 5 It is the linear laser triangulation principle diagram;

[0076] Figure 6 It is the spring measurement system measurement principle diagram;

[0077] Figure 7 It is the spring height measurement principle diagram.

[0078] Fig. 1 is a spring bearing table, 2, motion mechanism, 3, measuring head, 4, control system, 11, bearing table, 12, standard bar, 21, base structure, 22, translation mechanism, 23, sliding table, 24, drag chain mechanism, 25, panoramic camera fixing support, 31, first industrial camera, 32, second industrial camera, 33, panoramic camera, 34, line laser, 35, illuminating lamp, 41, control cabinet, 42, servo motor, 43, first travel switch, 44, second travel switch. DETAILED DESCRIPTION

[0079] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict, and the present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0080] To make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The description of the at least one exemplary embodiment is actually only illustrative, but not as any limitation on the present application and its application or use. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0081] Figure 1 is the first angle structure schematic diagram of a round spring height batch detection device based on visual laser scanning;

[0082] Figure 2 is the second angle structure schematic diagram of a round spring height batch detection device based on visual laser scanning;

[0083] A round spring height batch detection device based on visual laser scanning, comprising: a spring bearing table 1, a motion module, a detection module and a software module;

[0084] The spring bearing table 1 comprises a bearing table 11 and N standard bars 12; the bearing table 11 is used for placing the round spring to be detected; the N standard bars 12 are uniformly distributed on the spring bearing table 1 and are used for fitting the reference plane for measuring the height of the round spring;

[0085] The motion module comprises a motion mechanism 2 and a motion control system 4;

[0086] The motion mechanism 2 is arranged above the spring bearing table 1 and is used for realizing the translational motion mechanical device for laser scanning of the round spring to be detected;

[0087] Motion control system 4: for controlling the position and speed of the scanning translation movement of the motion mechanism 2;

[0088] Detection module: probe 3: for obtaining single detection three-dimensional data of the round spring to be detected;

[0089] Software module: for calculating three-dimensional coordinate data corresponding to the laser scanning line based on the single frame image transmitted by the probe based on the triangulation principle; and combining the motion information to form the spring three-dimensional point cloud data; processing the three-dimensional point cloud information to calculate the spring height.

[0090] Further: N standard rods 12: the preferred number of N is 9, and the 9 standard rods 12 are evenly distributed on the upper surface of the bearing table 11 in a 3x3 matrix, and the height of the standard rod 12 is slightly greater than the height of the round spring to be detected.

[0091] Further: the probe includes a first industrial camera 31, a second industrial camera 32, and a line laser 34;

[0092] Line laser 34: for projecting a laser line onto the upper surface of the round spring;

[0093] First industrial camera 31 and second industrial camera 32: for collecting the laser line image of the upper surface of the round spring during the entire detection process;

[0094] The first industrial camera 31 and the second industrial camera 32 are respectively arranged on both sides of the same horizontal line of the line laser 34.

[0095] Further: the motion mechanism 2 includes:

[0096] Infrastructure 21: for serving as a support body;

[0097] Translation mechanism 22: arranged above the infrastructure 21, driving the probe to move;

[0098] Drag chain mechanism 24: for accommodating various power lines and signal lines, and guiding the cable to move synchronously with the machine during scanning.

[0099] Figure 3 Motion control system schematic diagram of the round spring height batch detection device based on visual laser scanning;

[0100] Further: the motion control system 4 includes:

[0101] Servo motor 42: for driving the movement of the motion mechanism;

[0102] Servo driver: for driving the servo motor 42;

[0103] Motion controller: for controlling the servo driver;

[0104] First travel switch 43: arranged at one end of the translation mechanism 22, for measuring the limit;

[0105] Second travel switch 44: arranged at the other end of the translation mechanism 22, for measuring the limit.

[0106] Further: the detection module further comprises:

[0107] Panoramic camera 33: arranged above the measured spring bearing table 1, for obtaining the overall image of the measured spring;

[0108] Illumination lamp 35: for providing light source for the panoramic camera 33 to shoot.

[0109] Further: based on the principle of triangulation, the three-dimensional coordinate data corresponding to the laser scanning line is calculated; and combined with the motion information, the spring three-dimensional point cloud data is formed; the three-dimensional point cloud information is processed, and the spring height is calculated as follows:

[0110] The internal parameters and external parameters of the first industrial camera 31 and the second industrial camera 32 are calibrated;

[0111] According to the external parameters of the calibrated first industrial camera 31 and the second industrial camera 32, the conversion relationship between the camera coordinate system and the calibration coordinate system is obtained;

[0112] The plane equation of the laser plane in the calibration coordinate system is calibrated;

[0113] Based on the single-frame laser line image shot by the first industrial camera 31 and the second industrial camera 32, the pixel points of the laser line are extracted from the image, which are converted to the camera coordinate system by using the internal parameter matrix, and then converted to the calibration coordinate system by using the coordinate conversion, and the triangulation is formed with the laser plane, the three-dimensional point line corresponding to the single laser line of the first industrial camera 31 and the second industrial camera 32 is calculated, and the three-dimensional point lines corresponding to the two cameras are fused;

[0114] The three-dimensional point line is converted to the measurement coordinate system, combined with the moving vector, and the three-dimensional point cloud data of the spring and the standard rod 12 is formed;

[0115] The top surface data of the spring and the standard rod 12 is extracted by segmenting the spring three-dimensional point cloud data;

[0116] The reference plane is fitted by using the top surface data of the standard rod 12, the distance from the spring top surface data to the reference plane is calculated, the high points are selected according to the distance value in proportion, and the spring height is calculated by using the spring height algorithm.

[0117] Further: the process of calculating the spring height by using the spring height algorithm is as follows:

[0118] According to the acquired point cloud data, in The face projection data respectively extracts each circular spring center and the top surface data of the standard bar 12;

[0119] The top surface data of the standard bar 12 is used for plane fitting to obtain a reference surface for spring height measurement;

[0120] According to the spring center point and the spring top surface, a ring-shaped region is screened for detection points;

[0121] The distance from the detection point to the reference surface fitted by the top surface of the standard part is calculated;

[0122] The detection points are sorted in descending order according to the distance to the reference surface, and the top-ranked points are taken in proportion;

[0123] The top-ranked points are further plane fitted, and the circular ring range of the plane is determined;

[0124] The distance from the point on the plane in the circular ring to the nearest standard part is calculated;

[0125] The spring height is obtained by subtracting the nearest distance from the height of the standard bar 12.

[0126] Further, when the circular spring is a nested spring, the top surface data of the inner and outer springs is further segmented, and the inner and outer springs are divided by two ring-shaped regions based on the spring center, and the heights of the inner and outer springs are calculated based on the spring height algorithm.

[0127] Embodiment 1:

[0128] A circular spring height batch detection device based on visual laser scanning, comprising a spring bearing table 1, a motion module, a detection module and a software module;

[0129] The spring bearing table 1 comprises a bearing table 11 and 9 standard bars 12.

[0130] The bearing table 11 is used for placing and arranging the circular springs to be detected; the spring bearing table 1 is processed by high-strength cast steel, the flatness is strictly controlled within 0.05mm, and the stability of the spring placement reference is ensured; the surface is treated by anodic oxidation to improve wear resistance and corrosion resistance, and to adapt to the complex environment of the workshop.

[0131] The 9 standard bars 12 are evenly distributed on the upper surface of the bearing table 11 according to 3x3, and the height of the standard bar 12 is slightly greater than the height of the circular spring to be detected. The standard bar 12 is made of bearing steel material and is precisely ground, with a length accuracy of 0.02mm. The top surface data of the standard bar 12 is fitted by laser scanning to eliminate the influence of the slight deformation of the bearing table 11 on the measurement and ensure the uniformity of the height measurement reference.

[0132] Motion module: the motion module includes a motion mechanism 2 and a motion control system 4.

[0133] The motion mechanism 2 includes a base structure 21, a translation mechanism 22, a sliding table 23, and a drag chain mechanism 24.

[0134] The base structure 21 is a support body of the translation mechanism 22, which is processed and assembled by using section steel, and is subjected to failure treatment to eliminate internal stress and ensure overall rigidity. The parallelism error of the guide rail installed thereon is ≤0.03mm / m, which ensures the straightness of the scanning motion.

[0135] The translation mechanism 22 is arranged above the base structure 21 and drives the probe 3 to move. The translation mechanism 22 includes a synchronous belt and the sliding table 23. A servo motor 42 is used in cooperation with a high-precision synchronous belt transmission to drive the probe sliding table 23 to realize adjustable speed translation of 0-50mm / s, and the positioning accuracy reaches ±0.05mm, which meets the scanning requirements of different spring measurement processes.

[0136] The drag chain mechanism 24 is made of nylon material and is used for storing various power lines and signal lines to avoid signal interference. The service life of the drag chain mechanism 24 is more than 100,000 times. One end of the drag chain mechanism 24 is fixed to the probe sliding table 23 of the translation mechanism 22, and the cable is guided to move synchronously with the sliding table during the scanning process.

[0137] The motion control system, as shown in Figure 3 , includes a servo motor 42, a servo driver, a motion controller, a first travel switch 43, a second limit switch 44, and a PC, etc.

[0138] The servo motor 42 is used to control the movement of the translation structure 2.

[0139] The servo driver is used to receive the control signal of the motion controller to drive the servo motor 42.

[0140] The motion controller is used to control the servo driver. It communicates with the PC through USB, drives the motor through the servo driver, receives the limit switch signal, and controls the motion limit.

[0141] The first travel switch 43 and the second travel switch 44 are arranged at both ends of the translation mechanism 22 to transmit the switch signal to the motion controller to realize the measurement of the two end limits and protection.

[0142] PC machine: equipped with Intel Core i7 14 generation processor and 32GB memory, running Windows 11 system; communicate with the first industrial camera 31, the second industrial camera 32 and the panoramic camera 33 through the gigabit network card, the data transmission rate reaches 1Gbps, which meets the real-time image processing requirements; communicate with the servo motion controller through USB to realize motion control. The PC machine is arranged in the control cabinet 41;

[0143] The detection module includes three parts: the probe 3, the panoramic camera 33 and the illuminating lamp 35.

[0144] The probe 3 includes a line laser 34 and two industrial cameras, the first industrial camera 31 and the second industrial camera 32. The line laser 34 adopts a line laser with a wavelength of 420nm, which is projected onto the upper surface of the circular spring to form a laser line; the probe 3 is connected with the motion control system 4 through the probe connecting frame 34;

[0145] The first industrial camera 31 and the second industrial camera 32 are arranged on the two sides of the same horizontal line of the line laser 34 at a certain distance, which are used to collect the laser line image of the upper surface of the circular spring in the whole detection process, and communicate with the PC machine through the gigabit network cable to realize real-time uploading of the collected image.

[0146] The panoramic camera 33 is arranged above the measured spring bearing table 1 and is fixed on the panoramic camera fixing bracket 25, which is used to obtain the overall image of the measured spring and upload to the PC machine through the gigabit network cable.

[0147] The illuminating lamp 35 adopts LED industrial light source to provide illumination for the panoramic camera 33.

[0148] The software module includes two parts: the software interface sub-module and the detection algorithm sub-module.

[0149] The software interface sub-module:

[0150] The software interface sub-module, as shown in Figure 4 provides an operation interface for the user, which includes three parts: image display area, data display area and operation area.

[0151] The image display area displays the image collected by the panoramic camera 33, marks the spring contour and measurement points through image processing algorithm and coordinate conversion of the measurement system, nests the spring with different colors, and displays the spring number according to the spring position, which corresponds to the data in the data display area.

[0152] The data display area displays the number of each spring, the height of the outer spring, the height of the inner spring, the height difference between the inner and outer springs and other basic parameters, and highlights the unqualified items in red; supports data database recording, including detection time, operator ID and other additional information.

[0153] The operation area includes an open device button and a start measurement button;

[0154] “Open device” button: open the configuration camera and motion controller.

[0155] “Start measurement” button: start measurement;

[0156] “Stop” button: stop the motion if a problem is encountered during measurement.

[0157] “Interval” text box: set the time interval of image acquisition.

[0158] “Speed” text box: set the laser scanning movement speed.

[0159] “Distance” text box: set the laser scanning distance.

[0160] Detection algorithm submodule:

[0161] The detection algorithm submodule includes two parts: a three-dimensional point cloud acquisition algorithm and a spring height detection algorithm.

[0162] Three-dimensional point cloud acquisition algorithm:

[0163] The three-dimensional point cloud acquisition algorithm calibrates the internal and external parameters of the first industrial camera 31 and the second industrial camera 32; obtains the conversion relationship between the three camera coordinate systems and the calibration coordinate system according to the external parameters of the two cameras; calibrates the plane equation of the laser plane in the calibration coordinate system; calculates the three-dimensional point line corresponding to a single laser line of the first industrial camera 31 and the second industrial camera 32 according to triangulation, and fuses the corresponding three-dimensional point lines of the two cameras.

[0164] The internal and external parameters of the first industrial camera 31, the second industrial camera 32, and the panoramic camera 33 are calibrated.

[0165] The camera internal parameter calibration process is as follows:

[0166] The camera internal parameter calibration uses a planar checkerboard calibration board. By shooting multiple images of the calibration board in different poses, the internal and external parameters and distortion coefficients of the camera are solved based on the perspective projection principle and homography.

[0167] The calibration steps are briefly described as follows: data acquisition → calculation of homography matrix → solution of internal parameter matrix → calculation of external parameters based on the obtained initial internal parameters → solution of distortion coefficients → nonlinear optimization.

[0168] As shown in FIG. 5, the calibration coordinate system Oxyz, the camera coordinate system Oxyz, and the pixel coordinate system Ouv are set. Figure 5 ​​​; the intrinsic matrix is obtained through camera calibration Then, the conversion relationship between the camera coordinate system and the pixel coordinate system is as follows:

[0169] (1)

[0170] Wherein: is the pixel coordinate; is the coordinate of the space point in the camera coordinate system; is the depth of the space point along the optical axis direction to the camera optical center. is the camera intrinsic matrix, which satisfies

[0171] (2)

[0172] Wherein: , is the focal length (pixel unit) in the direction of , and is the principal point coordinate.

[0173] The calibration process of the camera extrinsic parameter is as follows:

[0174] Let a point in the calibration coordinate system be denoted as , the pose of the camera is described by the rotation matrix and the translation vector , the pixel coordinate is , that is:

[0175] (3)

[0176] Wherein: the camera pose is the extrinsic matrix of the camera.

[0177] The camera coordinate system after calibration is obtained, and the unit ray equation from the camera optical center to the space point is calculated. The three industrial cameras involved in the present application need to be calibrated for intrinsic and extrinsic parameters.

[0178] The basic principle of line laser triangulation is to project a laser line onto the surface of an object, and calculate the distance from the sensor to the measurement point through the principle of triangulation. The specific algorithm is as follows:

[0179] The calibration coordinate system is obtained through camera calibration and the conversion matrix of the camera coordinate system ; the pixel coordinate system of the camera is obtained relative to the intrinsic matrix of the camera coordinate system , which satisfies:

[0180] (4)

[0181] Wherein is the pixel coordinate; is the coordinate of the spatial point in the camera coordinate system; is the depth of the spatial point along the optical axis direction to the camera optical center, is the camera intrinsic matrix.

[0182] In order to facilitate the calculation of the depth information of the spatial point, the spatial coordinates are normalized, that is,

[0183] (5)

[0184] is the normalized coordinate, representing the direction vector from the camera optical center to the spatial point. The unitization processing is performed on the direction vector, denoted as , then:

[0185] (6)

[0186] is the unit ray vector from the camera optical center to the pixel point , which points to the real scene direction corresponding to the pixel point in the three-dimensional space.

[0187] Based on the captured multiple frames of laser line images, the pixel points are extracted from the laser lines, and the intrinsic matrix is used to convert them to the camera coordinate system to form a series of three-dimensional point clouds, and the specific process is as follows:

[0188] Light plane calibration: the camera captures multiple frames of laser line images, and uses image processing algorithm to extract pixel points from the laser lines, and uses intrinsic matrix to convert them to camera coordinate system to form a series of three-dimensional point clouds. The light plane in the camera coordinate system is fitted by using the point cloud data, and the laser plane parameters are obtained, and the specific principle is as follows:

[0189] The general equation of the plane in three-dimensional space can be expressed as:

[0190] (7)

[0191] where, is the plane normal vector, is the distance related parameter from the plane to the origin.

[0192] For the points in the point cloud data, ideally these points all satisfy the plane equation, but due to noise or error, in practice, the optimal plane parameters are found to minimize the sum of squared distances of all points to the plane.

[0193] The distance of the point to the plane :

[0194] (8)

[0195] The goal is to minimize the total sum of squared distances:

[0196] (9)

[0197] Since the denominator is irrelevant, minimizing is equivalent to minimizing the numerator:

[0198] (10)

[0199] To reduce parameter redundancy, normalize the plane equation, and further simplify by using centering. Calculate the centroid of the point cloud where:

[0200] (11)

[0201] Subtract the centroid coordinates from all point coordinates to get the centered coordinates:

[0202] (12)

[0203] The plane equation is now simplified to and the objective function becomes:

[0204] (13)

[0205] Arrange the centered point coordinates by row to form the matrix :

[0206] (14)

[0207] The plane parameters can be considered as a vector , so the objective function can be written as:

[0208] (15)

[0209] For this objective function, perform SVD decomposition on the matrix : The right singular vector corresponding to the smallest singular value is the solution vector of the overdetermined equation, which is the normal vector of the optimal fitting plane equation . Finally, combine the centroid to obtain the parameters of the plane equation : ​

[0210] (16)

[0211] Thus, the equation of the light plane in the camera coordinate system is obtained.

[0212] Establish a measurement system, such as... Figure 6 As shown, the calibration coordinate system is obtained through camera calibration. With the left camera coordinate system Right camera coordinate system Transformation matrix and The intrinsic parameter matrix of the left camera's image coordinate system relative to the camera coordinate system is obtained through camera calibration. The intrinsic parameter matrix of the right camera image coordinate system relative to the camera coordinate system The equations of the unit ray pointing from the optical center of the camera to the spatial point are calculated using formulas (4) to (6). The equations of the light plane are calculated using formulas (7) to (16). Based on the obtained laser plane parameters, the three-dimensional point cloud data corresponding to the laser points in a single image are obtained by simultaneously solving the unit ray equation and the light plane equation. .

[0213] Calibration of the coordinate system of the detection system With the calibration coordinate system Transformation matrix and laser scanning translation vector .

[0214] Transform the 3D point cloud data corresponding to the laser points in a single image into the coordinate system of the detection system. Below, and combined with the laser scanning translation vector and unit movement step size Obtain the collected 3D point cloud:

[0215] (17)

[0216] After the scan is completed, all the collected point cloud data are obtained, providing a data foundation for subsequent spring height calculation.

[0217] Figure 7 This is a schematic diagram of the spring height measurement principle;

[0218] Spring height measurement algorithm:

[0219] After acquiring 3D point cloud data, spring point cloud data is extracted by segmenting the 3D point cloud data. Top surface data of standard parts .

[0220] Based on the top surface data of the standard part Fitting the standard part top surface equation:

[0221] (18)

[0222] Base surface for spring height calculation

[0223] According to the segmented spring point cloud data In Face projection data respectively identify the center of each circular spring . With the center point as the center, divide the spring top surface annular area radius range , the plane distance of the point in the spring point cloud data to the center is:

[0224] (19)

[0225] If it satisfies , the point is identified as a point participating in height calculation.

[0226] Arrange the points in each annular area in descending order of height, take the first N% points, and form a point set .

[0227] Fit the plane to the point set , and set the fitting plane equation as:

[0228] (20)

[0229] The minimum distance of the annular area point to the base surface :

[0230] (21)

[0231] The final spring height calculation value is:

[0232] (22)

[0233] Where: is the height reference value of the standard part, is the correction coefficient.

[0234] In addition, for the case of nested inner and outer springs, the center of the spring needs to be divided into two annular areas for inner and outer springs, and the above method is used to calculate the height of the inner and outer springs respectively.

[0235] The working process of the circular spring height batch detection device based on visual laser scanning is as follows:

[0236] ​The measuring device is in situ: move the measuring translation mechanism 22 to the measurement starting position, and turn on the illuminating lamp 35;

[0237] Place the spring: place the spring to be measured in 4 columns on the bearing table 11, and the number of springs is less than the limit number of springs placed on the bearing table 11;

[0238] Place the bearing table 11: place the bearing table 11 on which the spring is placed into the position to be measured by using a forklift;

[0239] Take a panoramic image: when taking a panoramic image, the system automatically identifies the number and position of the spring, and if it does not match the preset number of placement (such as missing, misplacement), it immediately issues an audible and visual alarm to prompt the operator to adjust.

[0240] Set the measurement parameters: set the measurement interval, scanning speed, and measurement distance on the software interface (this operation can be performed in advance);

[0241] Start measuring: press the "start measuring" button to start measuring, and the device translation mechanism 22 carries the scanning probe 3 from the starting position to the scanning end position to collect spring scanning data;

[0242] Calculate the measurement results: after scanning, calculate the inner and outer spring heights according to the collected spring data, and display the calculation results in the data display area.

[0243] Generate a test report: after testing is completed, automatically generate a test report containing statistical data such as the number of qualified products, the number of unqualified products, and the qualified rate; unqualified springs are marked in the panoramic image, and are marked in red in the data display area for easy sorting.

[0244] Synchronize data to the cloud database: support querying historical test records through the unique identification of the spring (such as a two-dimensional code), and realize the whole life cycle traceability from production to scrap.

[0245] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or replace some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A batch detection device for the height of circular springs based on visual laser scanning, characterized in that: include: Spring support platform: includes a support platform and N standard bars; the support platform is used to place the round spring to be tested; N standard bars: evenly distributed on the spring support platform, used to fit the reference plane for measuring the height of the circular spring; Motion module: includes motion mechanism and motion control system; Motion mechanism: A mechanical device installed above the spring support platform to realize the translational motion of the circular spring under test for laser scanning; Motion control system: used to control the position and speed of the scanning translational motion of the motion mechanism; Detection module: Used to acquire single-segment 3D data of the spring to be tested; Software algorithm module: Based on the single-frame image transmitted by the detection module, it calculates the three-dimensional coordinate data corresponding to the laser scanning line based on the principle of triangulation, and combines it with motion information to form three-dimensional point cloud data of the spring; it processes the three-dimensional point cloud information to calculate the height of the spring; The process of calculating the three-dimensional coordinate data corresponding to the laser scanning line based on the principle of triangulation, and combining it with motion information to form three-dimensional point cloud data of the spring; the process of processing the three-dimensional point cloud information to calculate the height of the spring is as follows: The intrinsic and extrinsic parameters of the first and second industrial cameras were calibrated. The transformation relationship between the camera coordinate system and the calibration coordinate system is obtained by calibrating the external parameters of the first and second industrial cameras. Determine the plane equation of the laser surface in the calibration coordinate system; Based on single-frame laser line images captured by the first and second industrial cameras, the pixels of the laser lines are extracted from the images, and the intrinsic parameter matrix is ​​used to transform them into the camera coordinate system. After the coordinate system is transformed into the calibration coordinate system, triangulation is performed with the laser plane to calculate the three-dimensional point lines corresponding to the single laser lines of the first and second industrial cameras, and the three-dimensional point lines corresponding to the two cameras are fused. The three-dimensional points and lines are transformed into the measurement coordinate system and combined with the translation vector to form three-dimensional point cloud data of the spring and the standard rod. The top surface data of the spring and the standard rod are extracted by segmenting the three-dimensional point cloud data of the spring; Fit the reference plane to the top surface data of the standard bar, and calculate the distance from the top surface data of the spring to the reference plane; select high points according to the distance value and calculate the spring height using the spring height algorithm; The process of calculating the spring height using the spring height algorithm is as follows: Based on the acquired point cloud data, in The surface projection data are used to extract the center data of each circular spring and the top surface data of the standard bar. The reference plane for spring height measurement is obtained by using the top surface data of the standard bar for plane fitting. The points to be tested are selected based on the center point of the spring and the annular area detected on the top surface of the spring. Calculate the distance from the point to be tested to the reference plane fitted to the top surface of the standard part; The points to be tested are ranked from highest to lowest distance from the reference plane, and the points ranked first are selected proportionally. Perform plane fitting on the points that are ranked first, and determine the annular range of the plane; Find the shortest distance from a point on the inner plane of the annulus to the standard part; The height of the spring is obtained by subtracting the nearest distance from the height of the standard bar. The process of solving the plane equation in the camera coordinate system is as follows: For point cloud data Points point to plane Distance: (8) The goal is to minimize the total sum of squared distances: (9) Because of the denominator and Irrelevant, minimize Equivalent to minimizing the molecular part: (10) The plane equations are normalized and further simplified using centering, and the centroid of the point cloud is calculated. ,in: (11) Subtracting the centroid coordinates from all point coordinates yields the centered coordinates: (12) At this point, the plane equation simplifies to The objective function becomes: (13) Arrange the centered point coordinates in rows to form matrix : (14) Planar parameters Can be viewed as a vector Then the objective function can be written as: (15) For this objective function, for the matrix Perform SVD decomposition: The right singular vector corresponding to the minimum singular value is the solution vector of the overdetermined equation, which is also the normal vector of the optimal fitting plane equation. Finally, combined with the center of mass The parameters in the plane equation can be obtained. : (16) Thus, the equation of the light plane in the camera coordinate system is obtained.

2. The device for batch detection of circular spring height based on visual laser scanning according to claim 1, characterized in that: N standard bars: The number of N is 9. The 9 standard bars are evenly distributed on the upper surface of the support platform in a 3×3 matrix. At the same time, the height of the standard bars should match the height of the circular spring to be tested.

3. The device for batch detection of circular spring height based on visual laser scanning according to claim 1, characterized in that: The detection module includes a probe. The probe includes a first industrial camera, a second industrial camera, and a line laser; Line laser: used to project laser lines onto the upper surface of a circular spring; First industrial camera and second industrial camera: used to acquire laser line images of the upper surface of the circular spring throughout the entire inspection process; The first industrial camera and the second industrial camera are respectively positioned on opposite sides of the same horizontal line as the online laser.

4. The batch detection device for the height of circular springs based on visual laser scanning according to claim 3, characterized in that: The motion mechanism includes: Infrastructure: Used as the supporting structure; Translation mechanism: Located above the infrastructure, it drives the probe to move; Cable drag chain mechanism: used to store various power cables and signal cables. One end of the cable drag chain mechanism is fixed to the slide of the translation mechanism, and guides the cable to move synchronously with the slide during the scanning process.

5. The device for batch detection of circular spring height based on visual laser scanning according to claim 1, characterized in that: The detection module also includes: Panoramic camera: Set above the spring support platform under test, used to obtain an overall image of the spring under test; Lighting: Used to provide a light source for the panoramic camera to take pictures.

6. The device for batch detection of circular spring height based on visual laser scanning according to claim 1, characterized in that: It also includes further segmenting the top surface data of the inner and outer springs when the circular spring is a nested spring, dividing the inner and outer springs into two annular regions with the spring center as the dividing point, and calculating the height of the inner and outer springs respectively based on the spring height algorithm.

Images