A volume detection system integrating machine vision and contact sensors

Abstract

This invention discloses a volume detection system integrating machine vision and contact sensors, comprising a control system that coordinates a lifting component, a rotating component, a vision detection component, a contact detection component, and an integrated control unit. The control system is electrically connected to the motors, sensors, and cameras of each component, and is used to schedule the actions of each component, collect and process detection data, and perform data fusion and volume calculation. The vision detection component integrates 2D and 3D intelligent cameras for non-contact acquisition of the object's diameter and three-dimensional contour height information; the contact detection component uses a magnetostrictive displacement sensor for fixed-point height contact measurement of key points. The control system fuses and fits the global data from the vision system with the local high-precision data from the contact sensors in real time, ultimately accurately reconstructing the three-dimensional model of the measured object and calculating its volume.

Description

Technical Field

[0001] This invention belongs to the field of non-contact precision measurement and automated detection technology, specifically relating to a volume detection system that integrates machine vision and contact sensors, suitable for measuring the volume of irregular, volatile or high-precision objects. Background Technology

[0002] Currently, in the industrial manufacturing field, accurate measurement of the internal cavity volume of products is a crucial step in ensuring product quality and production consistency. Existing pure vision measurement technology can acquire rich shape information non-contactly and quickly. However, this technology is susceptible to the influence of object surface material, changes in ambient lighting, and complex background interference, leading to missing point cloud data or inaccurate contour extraction. For objects with special materials or complex shapes, measurement accuracy and stability are difficult to guarantee. Existing pure contact measurement technology offers extremely high measurement accuracy and is unaffected by the optical properties of the object surface. However, the measurement speed is slow and the efficiency is low, failing to meet the needs of online inspection. The mechanical probe's contact with the object surface can easily scratch the surface, posing a safety risk when measuring explosives. In recent years, there have been attempts to simply combine the two technologies, but these solutions often suffer from the problem of "information silos." The visual and contact sensing data have not achieved deep integration and collaborative control at the underlying level. Typically, measurements are taken independently and then the results are compared or simply stitched together. The measurement strategy is not dynamically adjusted according to the real-time characteristics of the object being measured, resulting in no substantial improvement in the overall system efficiency. Furthermore, the adaptive ability to objects with unknown characteristics is poor, and it cannot meet the measurement requirements of explosives. Summary of the Invention

[0003] To address the aforementioned problems, the purpose of this invention is to provide a volume detection system that integrates machine vision and contact sensors. This automated composite detection system combines mechanical motion, contact sensing, and machine vision, enabling efficient, non-contact automatic volume measurement and calculation for regular or irregular cavity products, significantly improving detection accuracy and automation level.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A volume detection system integrating machine vision and contact sensors includes a control system comprising a lifting assembly, a rotating assembly, a vision detection assembly, a contact detection assembly, and an integrated control unit that work together. The control system is electrically connected to the motors, sensors, and cameras of each assembly and is used to schedule the actions of each assembly, collect and process detection data, and complete data fusion and volume calculation.

[0005] Furthermore, the lifting assembly includes an explosion-proof servo motor A, a linear module, a motor mounting plate, a magnetostrictive displacement sensor A, a telescopic magnetic ring A, and a photoelectric sensor. The explosion-proof servo motor A is fixedly connected to the linear module via the motor mounting plate. The magnetostrictive displacement sensor A is fixedly connected to the motor mounting plate via a mounting bracket. The photoelectric sensor is mounted on the side of the linear module to detect the position of the motor mounting plate. The telescopic magnetic ring A cooperates with the magnetostrictive displacement sensor A. When the explosion-proof servo motor A drives the motor mounting plate to move vertically along the linear module, the telescopic magnetic ring A and the magnetostrictive displacement sensor A undergo relative displacement, and the displacement is fed back to the control system in real time.

[0006] The rotating assembly includes an explosion-proof servo motor B, a reducer, a swing arm frame, a swing arm mounting bracket, and a sensor bracket. The explosion-proof servo motor B is fixedly connected to the swing arm mounting bracket via the reducer, the swing arm frame is fixedly connected to the swing arm mounting bracket, and the sensor bracket is fixed to the swing arm frame via adjustable bolts, for mounting the vision inspection assembly. The explosion-proof servo motor B drives the swing arm frame to rotate, completing the position switching and positioning of the vision inspection assembly.

[0007] The visual inspection component includes a 2D intelligent camera subsystem and a 3D intelligent camera. The 3D intelligent camera is connected to the rotating component via a mounting bracket with adjustable bolts and acquires the 3D point cloud height data of the object under test using a line scanning method. The 2D intelligent camera subsystem includes a 2D intelligent camera, an explosion-proof housing, a light source, and a mounting bracket. The 2D intelligent camera is built into the explosion-proof housing, and the cable is connected to the control system via an explosion-proof connector. The 2D intelligent camera is connected to the rotating component via a mounting bracket with adjustable bolts, and the light sources are symmetrically distributed on both sides of the 2D intelligent camera. The 2D intelligent camera is used to acquire the 2D dimensions of the object under test.

[0008] The contact detection assembly includes a magnetostrictive displacement sensor B, a mounting bracket, a compression spring, guide rods, a telescopic magnetic ring B, a lower mounting frame, and a contact. The magnetostrictive displacement sensor B is fixedly connected to the mounting bracket. Two guide rods are mounted below the mounting bracket, and compression springs are sleeved on the guide rods. The telescopic magnetic ring B is mounted on the lower mounting frame, which is connected to the guide rods via a sliding bearing. The contact is mounted at the bottom of the lower mounting frame. After the contact contacts the object being measured, the compression spring is compressed, causing the lower mounting frame to displace relative to the magnetostrictive displacement sensor B. The telescopic magnetic ring B and the magnetostrictive displacement sensor B undergo relative displacement, and the displacement is fed back to the control system in real time to calculate the contact point height.

[0009] The control system performs data fusion as follows: the displacement data of the magnetostrictive displacement sensor A in the lifting assembly is recorded as a global height reference; the displacement data of the magnetostrictive displacement sensor B in the contact detection assembly is simultaneously acquired as an absolute height reference point; the two sets of displacement data are spatiotemporally synchronized and fitted, and combined with the two-dimensional dimensions and three-dimensional point cloud height data acquired by the visual detection assembly, a three-dimensional point cloud model of the object under test is generated through interpolation or surface reconstruction algorithms, and the volume of the object is calculated using integral or numerical calculation methods.

[0010] Compared with the prior art, the beneficial effects of the present invention are: (1) By deeply integrating the rapid, non-contact global perception capability of vision with the high precision and anti-interference fixed-point measurement capability of contact sensors, complementary advantages are achieved, and the overall performance of the system is improved. (2) Based on the preliminary visual analysis results, the system can intelligently judge and guide the contact probe to perform verification measurement on key and south-side areas, adapting to objects of different materials and shapes, and enhancing adaptability. (3) Under the premise of ensuring high precision, the overall measurement time for complex objects is significantly shortened, meeting the street photography requirements of industrial online detection. (4) Through optimized mechanism design and control logic, the functions of high-end measurement equipment are realized with relatively simple hardware combination. The present invention combines the high efficiency of non-contact measurement with the high precision and anti-interference capability of contact measurement, and is especially suitable for the volume detection of irregular, multi-material or high-precision industrial objects, and can better meet the measurement of the volume of explosive dangerous goods. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the overall structure of a volume detection system that integrates machine vision and contact sensors according to the present invention. Figure 2 This is a schematic diagram of the 2D smart camera structure described in this invention; Figure 3 This is a schematic diagram of the light source structure described in this invention; Figure 4 This is a schematic diagram of the 3D smart camera structure described in this invention; Figure 5 This is a schematic diagram of the contact detection component structure described in this invention; Figure 6 This is a schematic diagram of the rotating component structure described in this invention; Figure 7 This is a schematic diagram of the lifting component structure described in this invention. Detailed Implementation

[0012] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0013] See Figure 1-7A volume detection system integrating machine vision and contact sensors includes a control system comprising a lifting assembly 10, a rotating assembly 20, a vision detection assembly 30, a contact detection assembly 40, and an integrated control unit, all working in concert. The control system is electrically connected to the motors, sensors, and cameras of each assembly, and is used to schedule the actions of each assembly, collect and process detection data, and perform data fusion and volume calculation.

[0014] See Figure 7 The lifting assembly 10 includes an explosion-proof servo motor A 11, a linear module 12, a motor mounting plate, a magnetostrictive displacement sensor A 13, a telescopic magnetic ring A 14, and a photoelectric sensor 15. The explosion-proof servo motor A is fixedly connected to the linear module via the motor mounting plate. The magnetostrictive displacement sensor A is fixedly connected to the motor mounting plate via a mounting bracket. The photoelectric sensor 15 is mounted on the side of the linear module to detect the position of the motor mounting plate. The telescopic magnetic ring A cooperates with the magnetostrictive displacement sensor A. When the explosion-proof servo motor A drives the motor mounting plate to move vertically along the linear module, the telescopic magnetic ring and the magnetostrictive displacement sensor experience relative displacement, and the displacement is fed back to the control system in real time. When the lifting assembly is working, the explosion-proof servo motor A rotates, driving the motor mounting plate to move downwards. During the descent, the telescopic magnetic ring A and the magnetostrictive sensor A undergo relative displacement. This displacement is automatically recorded by the system and fitted with the displacement data uploaded by the magnetostrictive sensor B on the subsequent contact detection assembly to calculate the height of the product measurement point. When used in conjunction with the vision inspection assembly, the magnetostrictive sensor A is fitted with the camera detection height, and the system automatically calculates the height information of each point on the product, thereby automatically calculating the product volume.

[0015] See Figure 6 The rotating assembly 20 includes an explosion-proof servo motor B 21, a reducer 22, a swing arm frame 23, a swing arm mounting bracket 24, and a sensor bracket 25. The explosion-proof servo motor B 21 is fixedly connected to the swing arm mounting bracket via the reducer. The swing arm frame 23 is fixedly connected to the swing arm mounting bracket 24. The sensor bracket 25 is fixed to the swing arm frame 23 via adjustable bolts and is used to mount the vision inspection component. The explosion-proof servo motor B drives the swing arm frame to rotate, completing the position switching and positioning of the vision inspection component. When the rotating assembly is working, the lifting assembly 10 adjusts the rotating assembly 20 to a suitable height. The explosion-proof servo motor B rotates, driving the swing arm frame to rotate, rotating the vision inspection component and the contact inspection component to suitable positions for measurement by the respective inspection components.

[0016] See Figure 2-4The vision inspection component 30 includes a 2D smart camera subsystem and a 3D smart camera 31. The 3D smart camera 31 is connected to the rotating component via a mounting bracket with an adjustable bolt. During connection, the position of one end of the bolt is adjustable, allowing for camera position adjustment. The 3D smart camera is primarily used for height measurement, and its field of view 310 is as follows: Figure 4 As shown, a line scanning method is used to acquire the 3D point cloud height data of the object being measured. During scanning, the product rotates, and the rotating component adjusts its relative position to measure the height of all points within the product's end face. Combined with the diameter measured by the 2D intelligent camera 32, the internal volume of the product can be accurately measured. The 2D intelligent camera subsystem includes a 2D intelligent camera 32, an explosion-proof housing, a light source 33, and a mounting bracket 34. The 2D intelligent camera is built into the explosion-proof housing, and the cable connects to the control system via an explosion-proof connector 35. The 2D intelligent camera is connected to the rotating component via a mounting bracket with adjustable bolts. The light sources are symmetrically distributed on both sides of the 2D intelligent camera, each mounted at a 45° angle to the camera. The light sources are connected to the bracket with screws, and the bracket is connected to the rotating component via adjustable bolts. The 2D intelligent camera is used to acquire the two-dimensional dimensions of the object being measured. While the 2D intelligent camera itself is not explosion-proof, it meets explosion-proof standards when used with a dedicated explosion-proof housing.

[0017] See Figure 5 The contact detection assembly 40 includes a magnetostrictive displacement sensor B 41, a mounting bracket 42, a compression spring 43, a guide rod 44, a telescopic magnetic ring B 45, a lower mounting bracket 46, and a contact 47. The magnetostrictive displacement sensor B is fixedly connected to the mounting bracket. Two guide rods are mounted below the mounting bracket, and compression springs are fitted onto the guide rods. The telescopic magnetic ring B is mounted on the lower mounting bracket, which is connected to the guide rods via sliding bearings. The contact 47 is mounted at the bottom of the lower mounting bracket for contact with the product. After the contact 47 contacts the object being measured, the compression spring is compressed, causing the lower mounting bracket to displace relative to the magnetostrictive displacement sensor B. The telescopic magnetic ring B and the magnetostrictive displacement sensor B undergo relative displacement, and the displacement is fed back to the control system in real time. The contact point height is calculated using an internal algorithm.

[0018] The control system performs data fusion as follows: the displacement data of magnetostrictive displacement sensor A in the lifting assembly is recorded as the global height reference; the displacement data of magnetostrictive displacement sensor B in the contact detection assembly is acquired synchronously as the absolute height reference point; the two sets of displacement data are spatiotemporally synchronized and fitted, and combined with the two-dimensional dimensions and three-dimensional point cloud height data acquired by the vision detection assembly, a three-dimensional point cloud model of the object under test is generated through interpolation or surface reconstruction algorithms, and the volume of the object is calculated using integral or numerical calculation methods.

[0019] During operation: The system is powered on and initialized. The system power is connected, the control software is started, and the software will automatically read the zero-point parameters of each servo motor and sensor and perform a self-test. Figure 7 When the explosion-proof servo motor A in the lifting assembly shown is activated, it drives the linear module to raise the entire testing platform (including rotation and all testing components) to the preset "safe standby height" to avoid interference with the products on the workbench below. Figure 6 The explosion-proof servo motor B of the rotating component shown in the diagram rotates the swing arm to the "initial position" (which typically places the vision inspection component directly above the product or moves it out of the work area).

[0020] The work steps are as follows: 1. Install and roughly adjust the product: Place the product to be tested on the center positioning fixture of the worktable and fix it in place. The operator manually fine-tunes the lifting component through the control software so that the lens of the vision inspection component or the contact probe of the contact inspection component is roughly aligned with the center of the opening of the product cavity.

[0021] 2. Visual Inspection Component Implementation Steps (Non-Contact Scanning): During rotational positioning, the control software issues commands. The explosion-proof servo motor B of the rotation component precisely moves, driving the swing arm via a reducer to smoothly rotate the visual inspection component directly above the product. During precise height and focus determination, the software controls the explosion-proof servo motor A of the lifting component, driving the platform to slowly descend. Simultaneously, the platform's built-in magnetostrictive displacement sensor provides real-time height feedback. When the platform descends to the preset optimal visual scanning distance (triggered by a photoelectric sensor or a program-set position), the lifting stops. Adjustments are made manually or automatically based on ambient lighting and product surface characteristics. Figure 3 The brightness of the light source shown is adjusted to ensure uniform illumination. Refer to [reference needed] for instructions on adjusting the adjustable screws on the camera mounting bracket. Figure 2 and Figure 4 Fine-tune the angle and focal length of the 2D and 3D smart cameras to ensure clear images.

[0022] In the scanning and data acquisition process, Step 1 (2D Measurement): The 2D smart camera captures high-definition images of the product's end face under 45° illumination from both sides. The software uses image processing algorithms to accurately identify and calculate the two-dimensional dimensions of the inner cavity, such as its diameter and center position. Step 2 (3D Contour Scanning): Maintaining a constant height, the product's rotation mechanism is activated, causing the product to rotate at a uniform speed. Simultaneously, the 3D smart camera initiates line laser scanning, continuously scanning the inner wall of the cavity as the product rotates, acquiring depth (height) information for each point on the cavity's circumference, thereby constructing the three-dimensional point cloud contour of the entire cavity.

[0023] 3. Contact detection component implementation steps (contact verification): First, switch components. After visual inspection is completed, rotate the component again to isolate the visual inspection component and... Figure 5The contact detection component shown rotates to above the product. During the contact height measurement process, the entire platform and the contact detection component are slowly lowered by controlling the lifting mechanism. When the contact at the bottom of the component contacts the bottom of the product's inner cavity (or a specific depth reference surface requiring verification), the contact stops descending. However, because the lifting mechanism continues to descend slightly, the compression spring is compressed, pushing the lower mounting bracket connected to the guide rod to displace relative to the magnetostrictive displacement sensor body. This displacement causes a change in the position of the telescopic magnetic ring inside the sensor, and the amount of this displacement is captured by the sensor in real time with high precision.

[0024] 4. Data Fusion and Calculation: The system fits and calibrates the 3D contour data (relative height based on the camera coordinate system) obtained from visual inspection with one or more absolute height reference points obtained from contact inspection. This step eliminates potential overall height deviations in visual measurements, "anchoring" the relative coordinates of visual measurements to the absolute coordinate system. The software algorithm utilizes the calibrated, accurate 3D model of the cavity (containing the accurate height of each point and planar dimensions measured by a 2D camera) and employs integration or other numerical calculation methods to accurately calculate the true volume of the cavity. The results are automatically displayed on the operation page and can be transferred to the Production Management System (MES) to complete the entire inspection process.

[0025] The entire implementation process is an automated and sequential process of "preparation → visual scanning (acquiring contours) → contact verification (providing benchmarks) → data fusion (calibration calculations)". Through the precise collaboration of mechanical motion and the deep fusion of multi-sensor data, efficient and high-precision volume measurement is achieved.

Claims

1. A volume detection system integrating machine vision and contact sensors, characterized in that, The system includes a control system comprising a lifting assembly, a rotating assembly, a vision inspection assembly, a contact inspection assembly, and an integrated control unit that work together. The control system is electrically connected to the motors, sensors, and cameras of each assembly and is used to schedule the actions of each assembly, collect and process inspection data, and complete data fusion and volume calculation.

2. The system according to claim 1, characterized in that, The lifting assembly includes an explosion-proof servo motor A, a linear module, a motor mounting plate, a magnetostrictive displacement sensor A, a telescopic magnetic ring A, and a photoelectric sensor. The explosion-proof servo motor A is fixedly connected to the linear module via the motor mounting plate. The magnetostrictive displacement sensor A is fixedly connected to the motor mounting plate via a mounting bracket. The photoelectric sensor is mounted on the side of the linear module to detect the position of the motor mounting plate. The telescopic magnetic ring A cooperates with the magnetostrictive displacement sensor A. When the explosion-proof servo motor A drives the motor mounting plate to move vertically along the linear module, the telescopic magnetic ring A and the magnetostrictive displacement sensor A undergo relative displacement, and the displacement is fed back to the control system in real time.

3. The system according to claim 2, characterized in that, The rotating assembly includes an explosion-proof servo motor B, a reducer, a swing arm frame, a swing arm mounting bracket, and a sensor bracket. The explosion-proof servo motor B is fixedly connected to the swing arm mounting bracket via the reducer, the swing arm frame is fixedly connected to the swing arm mounting bracket, and the sensor bracket is fixed to the swing arm frame via adjustable bolts, for mounting the vision inspection assembly. The explosion-proof servo motor B drives the swing arm frame to rotate, completing the position switching and positioning of the vision inspection assembly.

4. The system according to claim 3, characterized in that, The visual inspection component includes a 2D intelligent camera subsystem and a 3D intelligent camera. The 3D intelligent camera is connected to the rotating component via a mounting bracket with adjustable bolts and acquires the 3D point cloud height data of the object under test using a line scanning method. The 2D intelligent camera subsystem includes a 2D intelligent camera, an explosion-proof housing, a light source, and a mounting bracket. The 2D intelligent camera is built into the explosion-proof housing, and the cable is connected to the control system via an explosion-proof connector. The 2D intelligent camera is connected to the rotating component via a mounting bracket with adjustable bolts, and the light sources are symmetrically distributed on both sides of the 2D intelligent camera. The 2D intelligent camera is used to acquire the 2D dimensions of the object under test.

5. The system according to claim 4, characterized in that, The contact detection assembly includes a magnetostrictive displacement sensor B, a mounting bracket, a compression spring, guide rods, a telescopic magnetic ring B, a lower mounting frame, and a contact. The magnetostrictive displacement sensor B is fixedly connected to the mounting bracket. Two guide rods are mounted below the mounting bracket, and compression springs are sleeved on the guide rods. The telescopic magnetic ring is mounted on the lower mounting frame, which is connected to the guide rods via a sliding bearing. The contact is mounted at the bottom of the lower mounting frame. After the contact contacts the object being measured, the compression spring is compressed, causing the lower mounting frame to displace relative to the magnetostrictive displacement sensor B. The telescopic magnetic ring B and the magnetostrictive displacement sensor B undergo relative displacement, and the displacement is fed back to the control system in real time to calculate the contact point height.

6. The system according to claim 5, characterized in that, The control system performs data fusion as follows: the displacement data of the magnetostrictive displacement sensor A in the lifting assembly is recorded as a global height reference; the displacement data of the magnetostrictive displacement sensor B in the contact detection assembly is simultaneously acquired as an absolute height reference point; the two sets of displacement data are spatiotemporally synchronized and fitted, and combined with the two-dimensional dimensions and three-dimensional point cloud height data acquired by the visual detection assembly, a three-dimensional point cloud model of the object under test is generated through interpolation or surface reconstruction algorithms, and the volume of the object is calculated using integral or numerical calculation methods.

Images