A microstructure light-based three-dimensional real-time adaptive measurement system and method for objects

Abstract

The application provides a kind of object three-dimensional real-time self-adaptive measurement system and measurement method based on microstructure light, system includes: multiple optical measurement units, equidistant circular ring array is installed on the ring mounting frame along the circumference and synchronous motion with the lifting adjusting mechanism;Multiple measurement posture acquisition units, reference posture monitoring unit, posture compensation unit, control unit, data fusion module, quality analysis module.According to the application, multiple optical measurement units arranged in circular ring array, realize the multi-view synchronous measurement of workpiece, avoid the time consumption and reference conversion error of traditional indexing measurement mode;Real-time posture monitoring and dynamic compensation mechanism effectively inhibits the measurement error caused by mechanical vibration, temperature drift and other factors, reduces the measurement error caused by posture;System realizes the automation, real-time and high-precision of three-dimensional measurement, improves the measurement efficiency and measurement accuracy.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional measurement technology, specifically to a real-time adaptive three-dimensional measurement system and method for objects based on microstructured light. Background Technology

[0002] With the continuous development of precision manufacturing and quality inspection technologies, high-precision dimensional measurement of three-dimensional objects has become an indispensable part of the industrial field. Currently, the industrial three-dimensional measurement field generally adopts structured light projection technology to achieve non-contact acquisition of the surface morphology of objects. This is done by projecting coded light patterns onto the measured surface and analyzing deformation fringes to calculate the three-dimensional coordinates. To achieve omnidirectional measurement of complex workpieces, existing technologies mostly use single-view measurement combined with precision indexing mechanisms. This involves rotating the worktable or moving the measured workpiece to complete multi-view data acquisition, and then using calibration target balls or feature point matching to stitch and fuse the point clouds from each viewpoint. In terms of measurement accuracy assurance, existing systems typically rely on factory calibration and periodic recalibration to eliminate system errors. Some high-end equipment is equipped with temperature compensation functions to suppress the effects of thermal drift.

[0003] However, in practical applications, multi-view time-sharing measurement methods result in low measurement efficiency. The introduction of a rotation mechanism increases system complexity and the risk of accumulated errors, and it is difficult to meet the cycle time requirements of real-time inspection on the production line. During the measurement process, equipment vibration, temperature fluctuations, and attitude drift caused by mechanical motion cannot be monitored and compensated in real time, resulting in dynamic measurement accuracy being significantly lower than static nominal accuracy. Multi-view data fusion relies on post-processing, splicing errors are difficult to control, and there is a lack of a unified real-time quality assessment mechanism, making it impossible to achieve adaptive optimization and closed-loop control of the measurement process. Therefore, this invention studies and designs a real-time adaptive 3D object measurement system and method based on microstructured light. Summary of the Invention

[0004] Therefore, the technical problem to be solved by the present invention is to overcome the above-mentioned defects, thereby providing a real-time adaptive three-dimensional measurement system and method for objects based on microstructured light.

[0005] To address the aforementioned problems, this invention provides a real-time adaptive 3D measurement system for objects based on microstructured light, comprising:

[0006] The workpiece placement platform is used to place the workpiece to be tested;

[0007] A lifting and adjusting mechanism is located above the work placement platform, and an annular mounting frame coaxial with the work placement platform is provided at the upper end of the lifting and adjusting mechanism.

[0008] Multiple optical measurement units are mounted on the ring mounting frame in an equidistant circular array along the circumference and move synchronously with the lifting and adjusting mechanism. The measuring optical axis of each optical measurement unit is oriented toward the central reference area of ​​the working platform, which is used to project structured light patterns onto the workpiece surface and acquire modulated structured light images.

[0009] Multiple attitude acquisition units are provided, each corresponding to one of the optical measurement units, and are rigidly mounted on the corresponding optical measurement units for real-time acquisition of the attitude data of the corresponding optical measurement units.

[0010] The reference attitude monitoring unit is fixedly installed on the work placement platform or equipment reference frame and is used to collect reference attitude data of the global reference position in real time.

[0011] An attitude compensation unit is connected to all the measured attitude acquisition units and the reference attitude monitoring unit respectively, and is used to generate dynamic compensation parameters for each optical measurement unit based on the difference between its corresponding measured attitude data and the reference attitude data.

[0012] The control unit is connected to the lifting adjustment mechanism, all the optical measurement units and the attitude compensation unit. It is used to generate a measurement strategy based on the three-dimensional design model of the workpiece to be measured, and to schedule the movement of the lifting adjustment mechanism, the working sequence and focal length adjustment of each optical measurement unit according to the measurement strategy. At the same time, it corrects the measurement action of each optical measurement unit according to the dynamic compensation parameters.

[0013] The data fusion module is connected to all the optical measurement units and the attitude compensation unit. It is used to receive the structured light images collected by each optical measurement unit and calculate multiple sets of original point cloud data. At the same time, it uses the corresponding dynamic compensation parameters to correct the attitude deviation of each set of original point cloud data. It then stitches and fuses the corrected multi-view point cloud data to generate a unified three-dimensional point cloud dataset based on the same reference coordinate system.

[0014] The quality analysis module is connected to the data fusion module and is used to calculate the geometric quality parameters of the workpiece based on the unified three-dimensional point cloud dataset.

[0015] Preferably, the optical measurement unit is a variable focal length optical measurement unit, including a focal length adaptive adjustment mechanism. The focal length adaptive adjustment mechanism changes the measurement focal length by physically sliding and adjusting the relative position of the optical elements. The focal length adaptive adjustment mechanism of each optical measurement unit can be adaptively slid along the radial direction of the annular mounting frame, thereby adapting to the measured surface at different radial positions and different heights on the workpiece surface.

[0016] Preferably, the variable focal length optical measurement unit includes:

[0017] Laser emitters are used to project structured light patterns onto the surface of a workpiece.

[0018] Image sensor, used to acquire structured light images modulated on the surface of a workpiece;

[0019] Optical lens groups are respectively disposed in the projection optical path of the laser emitter and the acquisition optical path of the image sensor;

[0020] The focal length adaptive adjustment mechanism includes a driving element and a sliding guide rail. The sliding guide rail is arranged along the radial direction of the annular mounting frame. At least one lens in the optical lens group is mounted on the sliding guide rail. The driving element drives the lens to slide along the radial direction according to the command of the control unit, so as to adjust the relative position of the optical element and change the measurement focal length.

[0021] Preferably, the attitude compensation unit includes:

[0022] The attitude difference calculation subunit is used to calculate, in real time, the angle difference between the corresponding measured attitude data and the reference attitude data in three orthogonal directions for each optical measurement unit.

[0023] The comprehensive error assessment subunit is used to weight and combine the angle differences in each direction according to preset weight parameters to obtain the differential attitude comprehensive error of the corresponding optical measurement unit, and to determine whether the differential attitude comprehensive error exceeds a preset threshold.

[0024] The compensation parameter generation subunit is used to generate dynamic compensation parameters containing the residual angles of each direction of the corresponding optical measurement unit when the differential attitude synthesis error exceeds the threshold.

[0025] Preferably, the compensation parameter generation subunit is further configured to determine the compensation time corresponding to the residual angle in each direction, and output the residual angle and the compensation time as compensation parameters; the compensation parameters gradually reduce their intensity in a linear or exponential manner within a preset subsequent time window.

[0026] Preferably, the data fusion module includes:

[0027] The point cloud processing subunit is used to process the structured light image acquired by each optical measurement unit into corresponding raw point cloud data.

[0028] The attitude correction subunit is connected to the attitude compensation unit and is used to perform rotation transformation on the coordinates of each point in each set of original point cloud data according to the dynamic compensation parameters of the corresponding optical measurement unit.

[0029] The multi-view stitching sub-unit is used to register, deduplicatize, and stitch together multiple sets of corrected point cloud data to generate a unified 3D point cloud dataset.

[0030] The confidence assessment subunit is used to assess the measurement confidence of each set of corrected point cloud data and send the confidence assessment results corresponding to each optical measurement unit to the control unit.

[0031] Preferably, the control unit includes:

[0032] The strategy planning module is used to identify the positioning surface, reference surface and detection surface based on the 3D design model of the workpiece to be measured. Combined with the layout position of the optical measurement units of the circular array, the measurement area of ​​each optical measurement unit is divided, and the initial measurement path and synchronous acquisition timing are generated.

[0033] The collaborative control module is connected to the strategy planning module for scheduling the movement of the lifting adjustment mechanism, the focal length adjustment and acquisition timing of each optical measurement unit, and real-time correction of the measurement actions of the corresponding optical measurement unit according to the dynamic compensation parameters.

[0034] The collaborative control module is configured to have at least two switchable control modes, including a first mode that executes measurement tasks according to the initial measurement path and the synchronous acquisition timing sequence, and a second mode that dynamically adjusts the measurement behavior based on the real-time quality parameters output by the quality analysis module or the confidence assessment results output by the data fusion module.

[0035] Preferably, the mode linkage logic of the collaborative control module is configured as follows:

[0036] If the real-time quality parameter exceeds a preset threshold or the confidence assessment result is lower than a preset standard during the operation of the first mode, the system will automatically switch from the first mode to the second mode.

[0037] In the second mode, the current sequence executed according to the strategy is paused, the target optical measurement unit corresponding to the abnormal area is located, and a new verification measurement is planned and executed for the abnormal area.

[0038] After the verification measurement is completed, switch back to the first mode to continue the remaining tasks based on the measurement results, or generate a new measurement task sequence based on the new measurement results.

[0039] Preferably, the quality analysis module is configured as follows:

[0040] The dimensional error analysis unit is used to calculate the deviation of data points in a unified 3D point cloud dataset from reference features in order to evaluate the flatness, parallelism, or spatial dimensional distance of the workpiece.

[0041] The surface topography deviation analysis unit is used to calculate the local curvature of the unified three-dimensional point cloud dataset and perform difference analysis with the three-dimensional design model to identify workpiece surface defects.

[0042] The report generation unit integrates the analysis results and generates a quantitative analysis report containing error statistics parameters, while simultaneously outputting real-time quality parameters to the control unit.

[0043] The present invention also provides a real-time adaptive three-dimensional measurement method for objects based on microstructured light, including the real-time adaptive three-dimensional measurement system for objects based on microstructured light as described in any of the preceding claims, comprising the following steps:

[0044] S1. Based on the three-dimensional design model of the workpiece to be measured, combined with the multi-optical measurement unit architecture of the circular array, a global measurement strategy and regional measurement tasks of each unit are generated.

[0045] S2. Control the lifting and adjusting mechanism to drive the ring mounting frame to move, and simultaneously schedule each optical measurement unit to collect structured light image data of the corresponding area of ​​the workpiece according to the measurement strategy, and simultaneously control the focal length adaptive adjustment mechanism of each optical measurement unit to complete the focal length adaptive adjustment along the radial direction;

[0046] S3. Synchronously collect the measurement attitude data of each optical measurement unit and the global reference attitude data of the reference attitude monitoring unit, calculate the attitude difference for each optical measurement unit, and generate the corresponding dynamic compensation parameters.

[0047] S4. The structured light images acquired by each optical measurement unit are respectively solved into raw point cloud data. The attitude deviation of each set of raw point cloud data is corrected by the corresponding dynamic compensation parameters. The corrected sets of point cloud data are stitched and fused to generate a unified three-dimensional point cloud dataset based on the same reference coordinate system.

[0048] S5. Calculate the geometric quality parameters of the workpiece based on the unified three-dimensional point cloud dataset and output a quality judgment report.

[0049] 1. This invention achieves multi-view synchronous measurement of workpieces through multiple optical measurement units arranged in a circular array, avoiding the time consumption and reference conversion errors of traditional indexing measurement methods; real-time attitude monitoring and dynamic compensation mechanisms effectively suppress measurement errors caused by factors such as mechanical vibration and temperature drift, reducing attitude-related measurement errors; automatic fusion of multi-viewpoint cloud data eliminates the inefficiency and inconsistency of traditional manual splicing, and the establishment of a unified coordinate system ensures the reference uniformity of full-size measurement; the system realizes automation, real-time measurement, and high precision of three-dimensional measurement, improving measurement efficiency and accuracy;

[0050] 2. This invention also constructs a complete closed loop from measurement planning to quality assessment; step S1 completes strategy planning before measurement, decomposing the global task into local tasks; step S2 executes multi-sensor collaborative measurement to achieve synchronization of motion control and optical measurement; step S3 monitors and compensates for attitude errors in real time to ensure the accuracy of measurement data; step S4 completes real-time fusion of multi-view data to establish a unified data foundation; step S5 performs quality analysis and report generation based on complete data; the data flow and control flow between each step are transmitted in real time through a high-speed industrial network to ensure the automation and real-time performance of the entire process;

[0051] 3. This invention also improves measurement efficiency by automating the entire 3D measurement process, seamlessly connecting multiple steps that traditionally require manual intervention; real-time attitude compensation and dynamic adjustment mechanisms ensure measurement accuracy and reduce dynamic errors; the establishment of a unified coordinate system eliminates the data silo problem of multi-view measurement and achieves complete digitization of the entire workpiece surface; automated quality analysis and report generation make measurement results immediately available, supporting online quality control and process optimization; it is suitable for rapid, high-precision 3D measurement of complex workpieces. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the system structure of the present invention;

[0053] Figure 2 This is a schematic diagram of the optical measurement unit structure and installation of the present invention;

[0054] Figure 3 This is a schematic diagram of the method flow of the present invention.

[0055] The reference numerals in the attached figures are as follows:

[0056] 1. Working platform; 2. Lifting and adjusting mechanism; 3. Optical lens assembly; 4. Drive element; 5. Sliding guide rail; 6. Housing; 7. Circular mounting bracket; 8. Sliding seat. Detailed Implementation

[0057] like Figure 1-2 As shown, this invention provides a real-time adaptive 3D measurement system for objects based on microstructured light, comprising:

[0058] The work platform 1 is used to place the workpiece to be measured to provide a stable and reliable measurement reference. A positioning fixture is installed in the center of the platform to fix workpieces of different shapes. The bottom of the platform is equipped with vibration isolation pads. The vibration isolation pads adopt a composite structure of natural rubber and metal layers to effectively isolate the influence of environmental vibration on the measurement.

[0059] The lifting adjustment mechanism 2 is located above the work placement platform 1. The upper end of the lifting adjustment mechanism 2 is provided with an annular mounting frame 7 coaxial with the work placement platform 1. The annular mounting frame 7 can be raised and lowered relative to the work placement platform 1 by an electric cylinder and a sliding rod. The inner diameter of the annular mounting frame 7 is determined according to the maximum outer diameter of the workpiece to be measured, so as to ensure that the optical measurement unit has sufficient measurement distance.

[0060] Multiple optical measurement units are mounted on the ring mounting frame 7 in an equidistant circular array along the circumference and move synchronously with the lifting and adjusting mechanism 2. The measuring optical axis of each optical measurement unit is oriented toward the central reference area of ​​the working platform 1, which is used to project structured light patterns onto the workpiece surface and acquire modulated structured light images.

[0061] Multiple attitude acquisition units, corresponding one-to-one with each of the optical measurement units, are rigidly mounted on the corresponding optical measurement units. They are used to acquire the attitude data of the corresponding optical measurement units in real time. The number of optical measurement units is determined according to the complexity of the workpiece, usually 4 to 8, and they are distributed at equal angles around the circumference of the ring mounting frame 7. For example, when 6 optical measurement units are configured, the included angle between adjacent units is 60 degrees. Each optical measurement unit is mounted on the ring mounting frame 7 through an adjustment mechanism, which includes a pitch adjustment seat and a yaw adjustment seat. The adjustment range is ±5 degrees and the adjustment accuracy is 0.01 degrees, used to accurately align with the central reference area.

[0062] The reference attitude monitoring unit is fixedly installed on the working platform 1 or the equipment reference frame. It is used to collect reference attitude data of the global reference position in real time. The reference attitude monitoring unit adopts a fiber optic gyroscope inertial navigation system. The specific model can be Honeywell HG1700. The installation position is selected at the geometric center of the equipment.

[0063] The attitude compensation unit is connected to all the measurement attitude acquisition units and the reference attitude monitoring unit respectively. It is used to generate dynamic compensation parameters for each optical measurement unit based on the difference between its corresponding measurement attitude data and the reference attitude data. The attitude compensation unit adopts a real-time controller, is equipped with an FPGA module, and uses industrial Ethernet for data connection to ensure real-time transmission of attitude data.

[0064] The control unit is signal-connected to the lifting adjustment mechanism 2, all the optical measurement units, and the attitude compensation unit. It is used to generate a measurement strategy based on the three-dimensional design model of the workpiece to be measured, and to schedule the movement of the lifting adjustment mechanism 2, the working sequence and focal length adjustment of each optical measurement unit according to the measurement strategy. At the same time, it corrects the measurement actions of each optical measurement unit according to the dynamic compensation parameters. The control unit adopts an industrial computer and runs measurement control software. The software integrates the Open CASCADE geometric kernel for parsing the three-dimensional design model.

[0065] The data fusion module is connected to all the optical measurement units and the attitude compensation unit. It receives structured light images acquired by each optical measurement unit and calculates multiple sets of raw point cloud data. Simultaneously, it corrects the attitude deviation of each set of raw point cloud data using the corresponding dynamic compensation parameters. The corrected multi-view point cloud data is then stitched and fused to generate a unified 3D point cloud dataset based on the same reference coordinate system. The data fusion module uses a workstation configured to accelerate point cloud processing. The software is developed based on the PCL point cloud library and integrates the ICP registration algorithm and the global optimization algorithm.

[0066] The quality analysis module is connected to the data fusion module and is used to calculate the geometric quality parameters of the workpiece based on the unified three-dimensional point cloud dataset. The quality analysis module runs detection software based on Geomagic Control to achieve comparative analysis with the design model.

[0067] Specifically, multiple optical measurement units arranged in a ring array achieve omnidirectional synchronous measurement of the workpiece. During the measurement process, the measurement posture acquisition unit on each optical measurement unit monitors the posture changes of the unit body in real time, the reference posture monitoring unit provides a globally stable posture reference, and the posture compensation unit calculates the difference between the two to generate dynamic compensation parameters. The control unit plans the measurement path according to the workpiece's three-dimensional design model and coordinates the synchronous acquisition of each unit. The data fusion module receives the images acquired by each unit, solves the point cloud, and uses the compensation parameters to correct the posture deviation, finally fusing them into a complete point cloud in a unified coordinate system. The quality analysis module performs geometric quality assessment based on the complete point cloud, and the entire process realizes a closed-loop workflow.

[0068] Specifically, this application achieves multi-view synchronous measurement of workpieces through multiple optical measurement units arranged in a circular array, avoiding the time consumption and reference conversion errors of traditional indexing measurement methods; the real-time attitude monitoring and dynamic compensation mechanism effectively suppresses measurement errors caused by factors such as mechanical vibration and temperature drift, reducing measurement errors caused by attitude; the automatic fusion of multi-view cloud data eliminates the inefficiency and inconsistency of traditional manual splicing, and the establishment of a unified coordinate system ensures the reference uniformity of full-size measurement; the system realizes the automation, real-time and high precision of three-dimensional measurement, improving measurement efficiency and measurement accuracy.

[0069] In some embodiments, the optical measurement unit is a variable focal length optical measurement unit, including a focal length adaptive adjustment mechanism. The focal length adaptive adjustment mechanism changes the measurement focal length by physically sliding and adjusting the relative position of the optical elements. The focal length adaptive adjustment mechanism of each optical measurement unit can be adaptively slid along the radial direction of the annular mounting frame 7, thereby adapting to the measured surface at different radial positions and different heights on the workpiece surface.

[0070] Specifically, the optical system of the variable focal length optical measurement unit adopts a dual telecentric optical path design, and the depth of field range can cover surface height changes of ±20mm through focal length adjustment; the sliding adjustment direction of the focal length adaptive adjustment mechanism is along the radial direction of the annular mounting frame 7, that is, along the extension line of the radius of the annular mounting frame 7; the optical measurement unit is mounted on the focal length adaptive adjustment mechanism; the control of the focal length adaptive adjustment mechanism adopts a linear grating ruler, and high-resolution feedback is achieved through fourth-harmonic subdivision. According to the target focal length parameters sent by the control unit, the position of the optical measurement unit is adjusted through a PID algorithm.

[0071] Specifically, during the measurement process, the control unit calculates the optimal focal length value required by each optical measurement unit based on the workpiece surface height and radial distance corresponding to the current measurement position; after receiving the instruction, the focal length adaptive adjustment mechanism drives the optical element to slide radially, changing the object distance and image distance relationship of the optical system to achieve continuous adjustment of the focal length; when the lifting adjustment mechanism 2 drives the ring mounting frame 7 to move up and down, each optical measurement unit adaptively adjusts the focal length according to the height of the current measurement layer to ensure that clear structured light images can be obtained on surfaces at different height positions.

[0072] In some embodiments, the variable-focus optical measurement unit includes: a laser emitter for projecting a structured light pattern onto the workpiece surface; an image sensor for acquiring a structured light image modulated by the workpiece surface; and optical lens groups 3, respectively disposed in the projection optical path of the laser emitter and the acquisition optical path of the image sensor. The focal length adaptive adjustment mechanism includes a driving element 4 and a sliding guide rail 5, the sliding guide rail 5 being arranged along the radial direction of the annular mounting frame 7, at least one lens in the optical lens group 3 being mounted on the sliding guide rail 5, and the driving element 4 driving the lens to slide along the radial direction according to the command of the control unit, so as to adjust the relative position of the optical elements and change the measurement focal length.

[0073] Specifically, the laser emitter uses a blue semiconductor laser, with a Powell prism at the front end to convert the Gaussian beam into a linear laser line with uniform intensity distribution. The laser emitter is mounted on an adjustment frame inside the housing 6. The image sensor uses an industrial CMOS camera with a double telecentric lens to ensure constant magnification at different object distances. The optical lens group 3 of the projection path includes at least a collimating lens, a beam expander, and a projection lens, all made of low-dispersion optical glass with an anti-reflection coating. The optical lens group 3 of the acquisition path includes an imaging objective and a filter. The driving element 4 can be an electric cylinder. The sliding guide rail 5 extends radially along the annular mounting frame 7, and a sliding seat 8 is provided on the sliding guide rail 5. The housing 6 of the optical measurement unit is mounted on the sliding seat 8. The front lens of the imaging objective is mounted on the edge of the sliding guide rail 5 away from the annular mounting frame 7. In the acquisition path, the lens group is made of two meniscus lenses bonded together, and the lenses are fixed to the sliding guide rail 5 by an aluminum alloy lens mount.

[0074] Specifically, the structured light stripes projected by the laser emitter are modulated by the workpiece surface, and the deformed stripes carrying surface height information are acquired by the image sensor; the focal length adaptive adjustment mechanism adjusts the equivalent focal length and working distance of the optical system by changing the position of the front lens of the imaging objective; when measuring surfaces at different radial positions, the drive element 4 moves the lens position according to the control command so that the imaging surface always coincides with the sensor target surface, ensuring the clarity of the acquired image.

[0075] In some embodiments, the attitude compensation unit includes: an attitude difference calculation subunit, used to calculate in real time the angle difference between the corresponding measured attitude data and the reference attitude data in three orthogonal directions for each optical measurement unit; a comprehensive error evaluation subunit, used to weight and combine the angle differences in each direction according to preset weight parameters to obtain the differential attitude comprehensive error of the corresponding optical measurement unit, and to determine whether the differential attitude comprehensive error exceeds a preset threshold; and a compensation parameter generation subunit, used to generate dynamic compensation parameters containing the residual angles in each direction of the corresponding optical measurement unit when the differential attitude comprehensive error exceeds the threshold.

[0076] Specifically, the three orthogonal directions of the attitude difference calculation subunit are defined as the pitch angle around the X-axis, the roll angle around the Y-axis, and the yaw angle around the Z-axis. The attitude difference calculation adopts a quaternion difference algorithm. First, the attitude data output by the measurement attitude acquisition unit and the reference attitude monitoring unit are converted into quaternion representations. After calculating the quaternion difference, it is converted into Euler angle difference. The calculation frequency is synchronized with the attitude data acquisition frequency to ensure real-time tracking of attitude changes. The preset weight parameters of the comprehensive error evaluation subunit are determined according to the degree of influence of the error in each direction on the measurement accuracy. For example, the weight coefficient of pitch angle and roll angle is 0.4, and the weight coefficient of yaw angle is 0.2, because the yaw angle error has a relatively small impact on radial measurement. The comprehensive error assessment subunit performs a weighted combination of the angle differences in each direction according to preset weight parameters to obtain the differential attitude comprehensive error of the corresponding optical measurement unit. The preset threshold is set according to the measurement accuracy requirements. For example, when the required measurement accuracy is 0.01mm, the corresponding angle threshold is set to 0.005 degrees. The comprehensive error assessment subunit is implemented using FPGA. The residual angle of the compensation parameter generation subunit is the angle difference in each direction at the current moment. The compensation parameters are output in the form of data packets, which include timestamp, unit identifier, residual angle values ​​in three directions, and comprehensive error value. The data packet format adopts JSON or binary structure and is sent to the data fusion module and control unit via UDP protocol.

[0077] Specifically, the attitude difference calculation subunit continuously receives data streams from the measurement attitude acquisition unit and the reference attitude monitoring unit, and calculates the attitude deviation of each optical measurement unit relative to the global reference in real time; the comprehensive error evaluation subunit quantifies the three-dimensional angle difference into a single comprehensive error index, and triggers a compensation decision after comparing it with a threshold; when the error exceeds the threshold, the compensation parameter generation subunit immediately generates compensation parameters containing specific residual angles for subsequent point cloud correction and measurement action adjustment.

[0078] In some implementations, the compensation parameter generation subunit is further configured to determine the compensation time corresponding to the residual angle in each direction, and output the residual angle and the compensation time as compensation parameters; the compensation parameters gradually reduce their intensity in a linear or exponential manner within a preset subsequent time window.

[0079] Specifically, the compensation time is determined based on the prediction of error change trends, using a first-order inertial filtering algorithm. The time constant is set from 50ms to 200ms according to the system response characteristics. When the comprehensive error change rate is less than the threshold, the compensation time is set to 100ms; when the error change rate is large, the compensation time is shortened to 50ms for rapid response; when the error tends to stabilize, the compensation time is extended to 200ms to avoid overshoot. The effect of the compensation parameter is weakened in a linear or exponential manner within a preset time window. The linear decay coefficient is 1 minus the ratio of the current time to the time window length, and the time window length is 500 milliseconds. The exponential decay function is the natural constant e raised to the power of (the negative ratio of the current time to the decay time constant), and the decay time constant is 200 milliseconds. The decay method is selected according to the error characteristics. The data structure of the compensation parameter is expanded to include a timestamp, unit identifier, residual angle values ​​in three directions, comprehensive error value, compensation time value, decay method identifier, and decay coefficient. The data packet length is fixed at 64 bytes to facilitate high-speed transmission and real-time processing.

[0080] Specifically, the compensation parameter generation subunit calculates the appropriate compensation time and attenuation method based on the error variation characteristics while generating the residual angle. In the point cloud correction stage, after receiving the compensation parameters, the attitude correction subunit applies the compensation within the compensation time window and gradually reduces the compensation weight according to the time progress. The linear attenuation method makes the compensation effect weaken uniformly, which is suitable for continuous slow drift. The exponential attenuation method makes the compensation effect start quickly and then slowly fade away, which is suitable for sudden vibration and shock. The attenuation mechanism ensures a smooth transition of the compensation effect, avoids the influence of parameter abrupt changes on the measurement results, and reduces the noise and fluctuations introduced by the compensation while ensuring the real-time compensation effect, making the measurement results smoother and more reliable.

[0081] In some embodiments, the data fusion module includes: a point cloud solving subunit, used to solve the structured light image acquired by each optical measurement unit into corresponding raw point cloud data; an attitude correction subunit, data-connected to the attitude compensation unit, used to perform rotation transformation on the coordinates of each point in each set of raw point cloud data according to the dynamic compensation parameters of the corresponding optical measurement unit; a multi-view stitching subunit, used to register, deduplicate, and stitch together the corrected sets of point cloud data to generate a unified three-dimensional point cloud dataset; and a confidence evaluation subunit, used to evaluate the measurement confidence of each set of corrected point cloud data and send the confidence evaluation results corresponding to each optical measurement unit to the control unit.

[0082] Specifically, point cloud computation employs a phase profilometry method based on triangulation principles. First, Fourier transform or phase shifting is performed on the acquired deformed fringe image to extract phase information. Then, phase unfolding is performed, and a quality-oriented path tracking algorithm or minimum norm method is used to recover the absolute phase. Finally, the phase values ​​are converted into three-dimensional coordinates according to the system calibration parameters. The computation process is accelerated in parallel on the GPU, with a single frame image computation time of less than 10ms and a point cloud density of 1 million points / frame. The rotation transformation uses a rigid body transformation matrix. A rotation matrix R is constructed based on the residual angles in the three directions of the compensation parameters. The rotation center is the projection center of the optical measurement unit. The corrected coordinates are equal to the rotation matrix R multiplied by (the difference between the original point coordinates and the projection center coordinates) plus the projection center coordinates. The rotation matrix R is constructed according to the Euler angles Δθx, Δθy, and Δθz in each direction in the ZYX order to ensure consistency with the coordinate system acquired by the attitude acquisition.

[0083] In some embodiments, the control unit includes:

[0084] The strategy planning module identifies positioning surfaces, reference surfaces, and detection surfaces based on the 3D design model of the workpiece under test. Combined with the placement of the optical measurement units in the circular array, it divides the measurement areas of each optical measurement unit and generates the initial measurement path and synchronous acquisition sequence. The 3D design model is input in STEP or IGES format and is accessed via Open. The CASCADE kernel analyzes geometric features; positioning surface identification is based on the area and normal vector direction of the surface, selecting the surface with the largest area and normal vector parallel to the Z-axis as the positioning surface; reference surface identification is based on geometric tolerance markings or manual specification; the detection surface is all surfaces whose geometric quality parameters need to be measured; the measurement area division adopts an algorithm based on visibility analysis, calculating the viewing angle and measurement distance of each optical measurement unit for each detection surface, and assigning the surface to the unit with the best visibility conditions; if a surface can be observed by multiple units, it is allocated according to the unit load balancing principle to ensure that the workload of each unit is uniform; the initial measurement path generation adopts the layer slicing method, dividing the workpiece into several measurement layers along the Z-axis direction, with the thickness of each layer set to 10mm to 20mm according to the depth of field range; the synchronous acquisition timing is determined according to the movement speed of the lifting adjustment mechanism 2 and the acquisition frame rate of the optical measurement unit, ensuring that acquisition is triggered after the mechanism moves to the target position, and the movement to the next position is completed immediately after acquisition.

[0085] Specifically, the collaborative control module is connected to the strategy planning module and is used to schedule the movement of the lifting adjustment mechanism 2, the focal length adjustment and acquisition timing of each optical measurement unit, and to correct the measurement actions of the corresponding optical measurement unit in real time according to the dynamic compensation parameters. The collaborative control module adopts a soft PLC architecture, runs a real-time Linux operating system, and has a control cycle of 1ms.

[0086] The collaborative control module is configured to have at least two switchable control modes, including a first mode that executes measurement tasks according to the initial measurement path and the synchronous acquisition timing sequence, and a second mode that dynamically adjusts the measurement behavior based on the real-time quality parameters output by the quality analysis module or the confidence assessment results output by the data fusion module.

[0087] Specifically, in the first mode, the collaborative control module strictly follows the timing table generated by the strategy planning module, the lifting and adjusting mechanism 2 moves along a preset path, the optical measurement unit triggers acquisition at a designated position, and the focal length adjustment mechanism adjusts according to the preset focal length curve. In the second mode, the collaborative control module receives geometric quality parameters from the quality analysis module and confidence assessment results from the data fusion module in real time. When an anomaly is detected, the current sequence is paused, and the local measurement task is replanned. The strategy planning module completes global planning before measurement, decomposes the measurement task of complex workpieces into local tasks of each unit, and optimizes the timing to improve efficiency. As the execution core, the collaborative control module ensures the efficiency and orderliness of the measurement process in the first mode. In the second mode, it achieves adaptive adjustment through real-time feedback. The switching between the two modes is triggered by preset conditions, ensuring that the system has both efficient conventional measurement capabilities and flexible adjustment capabilities to deal with anomalies.

[0088] Specifically, this application eliminates the subjectivity and uncertainty of manual planning through automatic planning based on a 3D design model; visibility analysis ensures reasonable allocation and complete coverage of the measurement area; layered slicing path planning matches the motion characteristics of the lifting and adjusting mechanism 2 to achieve efficient full-surface scanning; the dual-mode control architecture ensures measurement efficiency while giving the system the ability to handle anomalies and adaptive optimization, enabling the system to dynamically optimize the measurement strategy based on real-time quality feedback, thereby improving the reliability of measurement results and adaptability to complex situations.

[0089] In some implementations, the mode linkage logic of the collaborative control module is configured as follows: during the operation of the first mode, if the real-time quality parameter exceeds a preset threshold or the confidence assessment result is lower than a preset standard, the system automatically switches from the first mode to the second mode; the real-time quality parameter exceeding the preset threshold includes situations such as dimensional error exceeding the tolerance zone, geometric tolerance exceeding the standard, and surface defect size exceeding the allowable value; the confidence assessment result being lower than the preset standard refers to a confidence score lower than 0.6; the mode switching decision is determined by the collaborative control module in real time within the control cycle, with a switching delay of less than 2ms.

[0090] Specifically, in the second mode, the current sequence executed according to the strategy is paused, the target optical measurement unit corresponding to the abnormal area is located, and a verification measurement is re-planned and executed for the abnormal area. The abnormal area is located by analyzing the spatial distribution of quality parameters or the spatial location of low-confidence point clouds. The verification measurement planning includes: increasing the measurement density for the abnormal area and reducing the measurement layer thickness to 5 mm; adjusting the focal length and exposure parameters of the target optical measurement unit to optimize the imaging quality; and triggering adjacent optical measurement units to perform cross-validation measurements on the same area.

[0091] Specifically, after the verification measurement is completed, the system switches back to the first mode to continue executing the remaining tasks based on the measurement results, or generates a new measurement task sequence based on the new measurement results; if the verification measurement confirms that the anomaly is a real defect, the defect information is recorded and the remaining tasks continue to be executed; if the verification measurement indicates that the anomaly is a measurement error or noise, the previous measurement data is corrected based on the verification results, and then execution continues; if the verification measurement finds a systemic problem, such as anomalies in multiple areas, a new measurement task sequence is generated to increase the global measurement density or adjust the measurement parameters.

[0092] Specifically, the mode linkage logic constitutes the adaptive control closed loop of the measurement system; in the first mode, real-time monitoring continuously evaluates the measurement quality, and triggers mode switching immediately upon detecting an anomaly; in the second mode, anomaly handling distinguishes between real defects and measurement errors through local encrypted measurement and cross-validation, avoiding misjudgment; the verification results drive subsequent recovery strategies, ensuring the integrity of the measurement process and the accuracy of the data; the entire linkage logic is executed automatically in the real-time control system without manual intervention; the automatic mode switching mechanism realizes the monitoring of the measurement process and anomaly response, shortening the time for anomaly detection and response, and avoiding the continuous generation of invalid data; among them, the targeted verification measurement strategy improves the accuracy of anomaly judgment and reduces misjudgment and missed detection caused by measurement errors; the recovery mechanism ensures the continuity of the measurement process, enabling the completion of all measurement tasks and guaranteeing data quality even in the event of anomalies.

[0093] In some implementations, the quality analysis module is configured as: a dimensional error analysis unit, used to calculate the deviation of data points in a unified 3D point cloud dataset from reference features to evaluate the flatness, parallelism, or spatial dimensional distance of the workpiece; the reference features are extracted from the 3D design model and include basic geometric elements such as planes, cylindrical surfaces, and spheres; the flatness evaluation uses the minimum region method, extracting fitted planes from the point cloud data and calculating the difference between the maximum and minimum distances from each point to the plane, which is the flatness error; the parallelism evaluation extracts two fitted planes and calculates the angle between their normal vectors, or calculates the range of distance variation from each point on one plane to another plane; the spatial dimensional distance evaluation calculates the actual distance between two points, two lines, or two surfaces and compares it with the design dimensions to obtain the dimensional deviation; the evaluation results are presented in the form of a color deviation graph and a numerical table.

[0094] Specifically, the surface topography deviation analysis unit is used to calculate the local curvature of the unified 3D point cloud dataset and perform difference analysis with the 3D design model to identify workpiece surface defects. The local curvature calculation adopts a neighborhood-based least squares surface fitting method, with the neighborhood radius set from 0.5mm to 2mm, to calculate the principal curvature and Gaussian curvature of each point. The difference analysis calculates the normal deviation of each point after registering the measured point cloud with the design model. If the deviation exceeds 0.05mm and the area is greater than 1mm, the deviation is considered acceptable. 2 The area is identified as a defect; defect type identification includes: protrusions, depressions, scratches, pores, etc., classified according to curvature characteristics and deviation morphology.

[0095] Specifically, the report generation unit is used to integrate the analysis results and generate a quantitative analysis report containing error statistics parameters, while outputting real-time quality parameters to the control unit;

[0096] The quantitative analysis report includes: basic workpiece information, measurement conditions, geometric dimension inspection results, form and position tolerance evaluation results, surface defect statistics, and comprehensive quality judgment conclusions; it generates report formats such as PDF, Excel, and XML for easy archiving and system integration; real-time quality parameters are output in the form of structured data streams, with update frequency synchronized with the measurement frame rate, for real-time feedback control of the control unit.

[0097] Specifically, the dimensional error analysis unit reconstructs geometric features based on discrete point cloud data and quantifies the dimensional accuracy of the workpiece by comparing it with the design model; the surface topography deviation analysis unit identifies local anomalies on the surface through differential geometric analysis, realizing automatic detection and classification of defects; the report generation unit integrates the outputs of each analysis unit, generates a complete quality report in a standardized format, and extracts key quality indicators to feed back to the control system; the analysis process is automatically triggered after the measurement is completed, realizing an integrated process of measurement, analysis, and reporting.

[0098] like Figure 3As shown, the present invention also provides a real-time adaptive three-dimensional measurement method for objects based on microstructured light, including the real-time adaptive three-dimensional measurement system for objects based on microstructured light as described in any of the preceding claims, comprising the following steps:

[0099] S1. Based on the 3D design model of the workpiece to be measured, and combined with the multi-optical measurement unit architecture arranged in a circular array, a global measurement strategy and regional measurement tasks for each unit are generated. The 3D design model is imported into the control unit via Ethernet or USB interface, and the model format is STEP, IGES, or STL. The global measurement strategy includes measurement area division, measurement path planning, and acquisition parameter settings. The measurement area division adopts an automatic allocation algorithm based on visibility analysis, taking into account the viewing angle coverage, measurement distance, and resolution requirements of each optical measurement unit. The regional measurement tasks are issued to the local controller of each optical measurement unit in the form of a task list. The task list includes the coordinates of the measurement points, the focal length setting, the exposure parameters, and the trigger timing.

[0100] S2. Control the lifting and adjusting mechanism 2 to drive the annular mounting frame 7 to move, and synchronously schedule each optical measurement unit to acquire structured light image data of the corresponding area of ​​the workpiece according to the measurement strategy. Simultaneously control the focal length adaptive adjustment mechanism of each optical measurement unit to complete the adaptive focal length adjustment along the radial direction. The motion control of the lifting and adjusting mechanism 2 adopts an S-curve acceleration and deceleration plan, with a maximum speed of 50 mm / s and an acceleration of 100 mm / s. 2 This ensures smooth and shock-free movement; the acquisition trigger of each optical measurement unit is synchronized with the position of the lifting and adjusting mechanism 2, with a position synchronization accuracy of ±0.1mm; the adaptive focal length adjustment calculates the target focal length by looking up a table or interpolating based on the radial distance and height value of the current measurement point, with an adjustment response time of less than 50ms, ensuring that the focal length is locked before image acquisition;

[0101] S3. Synchronously acquire the measurement attitude data of each optical measurement unit and the global reference attitude data of the reference attitude monitoring unit, calculate the attitude difference for each optical measurement unit, and generate corresponding dynamic compensation parameters; attitude data acquisition and image acquisition are triggered synchronously, with a timestamp alignment accuracy of 1ms; attitude difference calculation is completed in real time in the FPGA, with a calculation delay of less than 1μs; dynamic compensation parameters include residual angles in three directions, compensation time, and attenuation method. After the parameters are generated, they are immediately sent to the data fusion module and control unit for subsequent point cloud correction and measurement control adjustment;

[0102] S4. The structured light images acquired by each optical measurement unit are respectively processed into raw point cloud data. The attitude deviation of each set of raw point cloud data is corrected using the corresponding dynamic compensation parameters. The corrected sets of point cloud data are then stitched together to generate a unified 3D point cloud dataset based on the same reference coordinate system. The structured light image processing is performed in parallel on the GPU cluster, using CUDA acceleration, with a single frame processing time of less than 10ms. The attitude deviation correction uses rigid body transformation, and the transformation matrix is ​​constructed in real time according to the dynamic compensation parameters. The stitching of multiple sets of point clouds uses a global registration algorithm, with a registration error of less than 0.01mm. The fused unified 3D point cloud dataset contains complete surface information of the entire measurement area.

[0103] S5. Calculate the geometric quality parameters of the workpiece based on the unified 3D point cloud dataset and output a quality judgment report; the geometric quality parameter calculation includes dimensional error, geometric tolerance and surface defects, and the calculation is based on the PCL point cloud library; the quality judgment report is automatically generated according to the preset tolerance standard, including qualified / unqualified judgment, out-of-tolerance statistics and defect location marking; the report output format supports PDF, Excel and database import, and key quality parameters are displayed on the human-machine interface in real time.

[0104] Specifically, this application constructs a complete closed loop from measurement planning to quality assessment; step S1 completes the strategy planning before measurement, decomposing the global task into local tasks; step S2 executes multi-sensor collaborative measurement to achieve synchronization of motion control and optical measurement; step S3 monitors and compensates for attitude errors in real time to ensure the accuracy of measurement data; step S4 completes the real-time fusion of multi-view data to establish a unified data foundation; step S5 performs quality analysis and report generation based on complete data; the data flow and control flow between each step are transmitted in real time through a high-speed industrial network to ensure the automation and real-time performance of the entire process.

[0105] This application achieves full automation of the 3D measurement process, seamlessly connecting multiple steps that traditionally require manual intervention, thereby improving measurement efficiency; real-time attitude compensation and dynamic adjustment mechanisms ensure measurement accuracy and reduce dynamic errors; the establishment of a unified coordinate system eliminates the data silo problem of multi-view measurement and realizes complete digitization of the entire surface of the workpiece; automated quality analysis and report generation make the measurement results immediately available, supporting online quality control and process optimization; suitable for rapid and high-precision 3D measurement of complex workpieces.

[0106] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention. The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the protection scope of the present invention.

Claims

1. A real-time adaptive three-dimensional measurement system for objects based on microstructured light, characterized in that, include: The workpiece placement platform is used to place the workpiece to be tested; A lifting and adjusting mechanism is located above the work placement platform, and an annular mounting frame coaxial with the work placement platform is provided at the upper end of the lifting and adjusting mechanism. Multiple optical measurement units are mounted on the ring mounting frame in an equidistant circular array along the circumference and move synchronously with the lifting and adjusting mechanism. The measuring optical axis of each optical measurement unit is oriented toward the central reference area of ​​the working platform, which is used to project structured light patterns onto the workpiece surface and acquire modulated structured light images. Multiple attitude acquisition units are provided, each corresponding to one of the optical measurement units, and are rigidly mounted on the corresponding optical measurement units for real-time acquisition of the attitude data of the corresponding optical measurement units. The reference attitude monitoring unit is fixedly installed on the work placement platform or equipment reference frame and is used to collect reference attitude data of the global reference position in real time. An attitude compensation unit is connected to all the measured attitude acquisition units and the reference attitude monitoring unit respectively, and is used to generate dynamic compensation parameters for each optical measurement unit based on the difference between its corresponding measured attitude data and the reference attitude data. The control unit is connected to the lifting adjustment mechanism, all the optical measurement units and the attitude compensation unit. It is used to generate a measurement strategy based on the three-dimensional design model of the workpiece to be measured, and to schedule the movement of the lifting adjustment mechanism, the working sequence and focal length adjustment of each optical measurement unit according to the measurement strategy. At the same time, it corrects the measurement action of each optical measurement unit according to the dynamic compensation parameters. The data fusion module is connected to all the optical measurement units and the attitude compensation unit. It is used to receive the structured light images collected by each optical measurement unit and calculate multiple sets of original point cloud data. At the same time, it uses the corresponding dynamic compensation parameters to correct the attitude deviation of each set of original point cloud data. It then stitches and fuses the corrected multi-view point cloud data to generate a unified three-dimensional point cloud dataset based on the same reference coordinate system. The quality analysis module is connected to the data fusion module and is used to calculate the geometric quality parameters of the workpiece based on the unified three-dimensional point cloud dataset. The attitude compensation unit includes: The attitude difference calculation subunit is used to calculate, in real time, the angle difference between the corresponding measured attitude data and the reference attitude data in three orthogonal directions for each optical measurement unit. The comprehensive error assessment subunit is used to weight and combine the angle differences in each direction according to preset weight parameters to obtain the differential attitude comprehensive error of the corresponding optical measurement unit, and to determine whether the differential attitude comprehensive error exceeds a preset threshold. The compensation parameter generation subunit is used to generate dynamic compensation parameters containing the residual angles of each direction of the corresponding optical measurement unit when the differential attitude synthesis error exceeds the threshold. The compensation parameter generation subunit is also configured to determine the compensation time corresponding to the residual angle in each direction, and output the residual angle and compensation time as compensation parameters; the compensation parameters gradually weaken their effect in a linear or exponential manner within a preset subsequent time window.

2. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 1, characterized in that, The optical measurement unit is a variable focal length optical measurement unit, including a focal length adaptive adjustment mechanism. The focal length adaptive adjustment mechanism changes the measurement focal length by physically sliding and adjusting the relative position of the optical elements. The focal length adaptive adjustment mechanism of each optical measurement unit can be adaptively slid and adjusted along the radial direction of the annular mounting frame, thereby adapting to the measured surface at different radial positions and different heights on the workpiece surface.

3. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 2, characterized in that, The variable focal length optical measurement unit includes: Laser emitters are used to project structured light patterns onto the surface of a workpiece. Image sensor, used to acquire structured light images modulated on the surface of a workpiece; Optical lens groups are respectively disposed in the projection optical path of the laser emitter and the acquisition optical path of the image sensor; The focal length adaptive adjustment mechanism includes a driving element and a sliding guide rail. The sliding guide rail is arranged along the radial direction of the annular mounting frame. At least one lens in the optical lens group is mounted on the sliding guide rail. The driving element drives the lens to slide along the radial direction according to the command of the control unit, so as to adjust the relative position of the optical element and change the measurement focal length.

4. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 1, characterized in that, The data fusion module includes: The point cloud processing subunit is used to process the structured light image acquired by each optical measurement unit into corresponding raw point cloud data. The attitude correction subunit is connected to the attitude compensation unit and is used to perform rotation transformation on the coordinates of each point in each set of original point cloud data according to the dynamic compensation parameters of the corresponding optical measurement unit. The multi-view stitching sub-unit is used to register, deduplicatize, and stitch together multiple sets of corrected point cloud data to generate a unified 3D point cloud dataset. The confidence assessment subunit is used to assess the measurement confidence of each set of corrected point cloud data and send the confidence assessment results corresponding to each optical measurement unit to the control unit.

5. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 1, characterized in that, The control unit includes: The strategy planning module is used to identify the positioning surface, reference surface and detection surface based on the 3D design model of the workpiece to be measured. Combined with the placement of the optical measurement units of the circular array, the measurement area of ​​each optical measurement unit is divided, and the initial measurement path and synchronous acquisition timing are generated. The collaborative control module is connected to the strategy planning module for scheduling the movement of the lifting adjustment mechanism, the focal length adjustment and acquisition timing of each optical measurement unit, and real-time correction of the measurement actions of the corresponding optical measurement unit according to the dynamic compensation parameters. The collaborative control module is configured to have at least two switchable control modes, including a first mode that executes measurement tasks according to the initial measurement path and the synchronous acquisition timing sequence, and a second mode that dynamically adjusts the measurement behavior based on the real-time quality parameters output by the quality analysis module or the confidence assessment results output by the data fusion module.

6. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 5, characterized in that, The mode linkage logic of the collaborative control module is configured as follows: If the real-time quality parameter exceeds a preset threshold or the confidence assessment result is lower than a preset standard during the operation of the first mode, the system will automatically switch from the first mode to the second mode. In the second mode, the current sequence executed according to the strategy is paused, the target optical measurement unit corresponding to the abnormal area is located, and a new verification measurement is planned and executed for the abnormal area. After the verification measurement is completed, switch back to the first mode to continue the remaining tasks based on the measurement results, or generate a new measurement task sequence based on the new measurement results.

7. The real-time adaptive three-dimensional measurement system for objects based on microstructured light according to claim 1, characterized in that, The quality analysis module is configured as follows: The dimensional error analysis unit is used to calculate the deviation of data points in a unified 3D point cloud dataset from reference features in order to evaluate the flatness, parallelism, or spatial dimensional distance of the workpiece. The surface topography deviation analysis unit is used to calculate the local curvature of the unified three-dimensional point cloud dataset and perform difference analysis with the three-dimensional design model to identify workpiece surface defects. The report generation unit integrates the analysis results and generates a quantitative analysis report containing error statistics parameters, while simultaneously outputting real-time quality parameters to the control unit.

8. A real-time adaptive three-dimensional measurement method for objects based on microstructured light, characterized in that: The real-time adaptive 3D measurement system for objects based on microstructured light, as described in any one of claims 1-7, includes the following steps: S1. Based on the three-dimensional design model of the workpiece to be measured, combined with the multi-optical measurement unit architecture of the circular array, a global measurement strategy and regional measurement tasks of each unit are generated. S2. Control the lifting and adjusting mechanism to drive the ring mounting frame to move, and simultaneously schedule each optical measurement unit to collect structured light image data of the corresponding area of ​​the workpiece according to the measurement strategy, and simultaneously control the focal length adaptive adjustment mechanism of each optical measurement unit to complete the focal length adaptive adjustment along the radial direction; S3. Synchronously collect the measurement attitude data of each optical measurement unit and the global reference attitude data of the reference attitude monitoring unit, calculate the attitude difference for each optical measurement unit, and generate the corresponding dynamic compensation parameters. S4. The structured light images acquired by each optical measurement unit are respectively solved into raw point cloud data. The attitude deviation of each set of raw point cloud data is corrected by the corresponding dynamic compensation parameters. The corrected sets of point cloud data are stitched and fused to generate a unified three-dimensional point cloud dataset based on the same reference coordinate system. S5. Calculate the geometric quality parameters of the workpiece based on the unified three-dimensional point cloud dataset and output a quality judgment report.

Images