Timing-synchronized high-frequency line structured light dynamic measurement method, system and calibration platform

Abstract

The present invention discloses a timing-synchronized high-frequency line structured light dynamic measurement method, system and calibration platform, the method includes S100: emitting line laser and detecting light to the surface of an object, collecting the laser line image after being reflected by the surface of the object, and extracting the laser center line of the laser line image to obtain the laser center line position; S200: receiving an external second pulse time signal to perform second clock correction, and at the same time, performing high-precision time subdivision on the corresponding second clock to unify the time base of each module, realizing time synchronization through sensor synchronization control, and recording the working time of the sensor; S300: calibrating the laser center line position to obtain the measurement data of the line structured light sensor; S400: packaging the measurement data with the working time of the sensor to obtain the line structured light three-dimensional measurement data with high-precision time information, and providing an accurate data basis for multi-sensor data fusion. The present invention realizes high-frequency and high-precision three-dimensional measurement with time information.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical sensors, and more specifically, relates to a timing synchronized high-frequency line structured light dynamic measurement method, system and calibration platform. Background Art

[0002] Line structured light sensors have attracted attention due to their wide applications in three-dimensional measurement, machine vision and other fields. In dynamic measurement environments or when multiple sensors work together, the measurement accuracy of the system depends largely on the time synchronization accuracy of the collected data. In traditional methods, the accuracy of time synchronization is limited by the stability and accuracy of the local clock source, which is often difficult to meet the requirements in the field of high-precision measurement. Multi-sensor synchronization control is a core part of mobile measurement technology, and it plays a decisive role in the accuracy and integrity of data. When the measurement system integrates multiple sensor devices such as GNSS, INS, image sensors, and line structured light, it is necessary to ensure that these devices work together on the basis of time synchronization to maintain the accuracy and consistency of the measurement data.

[0003] The core concept of time synchronization is to ensure that multiple sensors work within the same time frame so that data can be accurately correlated and fused. Time synchronization usually involves the establishment of a high-precision time reference, time transfer, multi-sensor synchronization control methods, and the calibration and compensation of the sensor's own delay. By studying the factors that affect the time transfer error from the time reference to the sensor, it is found that the sensor's own delay is the key point affecting the time uniformity of the measurement system. The time reference is usually established by using the precise time signal of the Global Positioning System (GPS). GPS satellites carry high-precision atomic clocks that can provide very accurate time signals. Typically, this signal is provided in the form of 1PPS (one pulse per second), representing the beginning of each second. The time signal received from GPS can be regarded as a global time reference. The transmission of time signals from the GPS receiver to each sensor involves the transmission delay of the signal in the circuit. Indicates the time when the signal is sent from the GPS receiver. represents the time when the sensor receives the signal, the transmission delay can be expressed as The error in time transmission includes the propagation delay of the signal in the cable and circuit, the processing delay in the circuit, etc. The propagation delay of the signal in the cable is recorded as , the circuit processing delay is recorded as , the total time transfer error can be expressed as After determining the delay in time transfer, the system can take compensating measures for each sensor to ensure that all data is tagged with the same time reference. is the time stamp of the sensor recording data. The corrected time stamp can be expressed as Through the above steps, high-precision time synchronization between multiple sensors can be achieved. This is essential to ensure the spatiotemporal consistency of data collected in mobile measurement systems, especially when processing and analyzing these data to obtain accurate measurement results.

[0004] The data collection of line structured light sensors in dynamic environments needs to be synchronized with precise time information to ensure the accuracy and reliability of the data. Existing synchronous timing technology often cannot meet the needs of high-dynamic and high-precision applications. The actual working time of the sensor Difficult to record accurately: Different sensors may have different response times, which may change due to many factors, such as the hardware characteristics of the sensor, temperature changes, power supply fluctuations, circuit design, etc. The sensor response time has great uncertainty. Time transmission errors are difficult to quantify and count. Signal transmission in cables and circuits will cause delays. These delays may vary due to environmental and hardware differences. Cable length and type, circuit board design, electrical noise, etc. It is difficult to calculate the specific time transmission error value through an accurate model. Secondly, the performance of electronic components may be affected by temperature, humidity and other environmental factors. When working for a long time, environmental conditions (such as temperature, humidity), component aging, unstable power supply, etc. will also interfere with the unification of time signals. However, existing structured light sensors cannot accurately record the absolute time of their own operation, but can only record relative time. It is difficult to unify the measurement data and to fuse it with other sensor data. It is impossible to achieve high-frequency and high-precision measurement in motion. Summary of the invention

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a timing synchronized high-frequency line structured light dynamic measurement method, system and calibration platform, which establishes a high-precision time reference based on sensors through a synchronous core processing method, and combines PPS pulses with high-stability crystal oscillators to not only maintain the operation of the time reference, but also accurately record the working time of the sensor with the crystal oscillator frequency as the resolution, thereby realizing the time synchronization of multiple sensors. At the same time, a special form of calibration object is designed to have calibration features that can be quickly and accurately extracted, and the image space coordinates are converted into object space coordinates through the calibration object constrained by virtual points and the extraction method, thereby obtaining the measurement data of the line structured light sensor, thereby realizing high-frequency and high-precision measurement in motion.

[0006] To achieve the above object, according to one aspect of the present invention, a timing synchronized high-frequency line structured light dynamic measurement method is proposed, comprising:

[0007] S100: emitting a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collecting a laser line image after being reflected by the surface of the object, and extracting a laser center line of the laser line image to obtain a laser center line position;

[0008] S200: Receive external second pulse time signal to calibrate the second clock, and at the same time, perform high-precision time subdivision on the corresponding second clock to unify the time base of each module, realize time synchronization through sensor synchronization control, and record the working time of the sensor;

[0009] S300: Calibrate the laser center line position, and convert the image space coordinates into object space coordinates through a virtual point constrained calibration object and an extraction method, so as to obtain measurement data of the line structured light sensor;

[0010] S400: Packing the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information, providing an accurate data basis for multi-sensor data fusion.

[0011] Furthermore, the time synchronization in step S200 includes:

[0012] S201: define a differential crystal pulse counter and a second-level time counter, and receive an external pulse time signal to initialize the second-level time counter;

[0013] S202: Count the external pulse time signal and continuously monitor the second time counter signal. When a rising signal of the second time counter is detected, the second time counter increases by one step and resets the differential crystal oscillator pulse counter at the same time. The count values ​​of the differential crystal oscillator pulse counter and the second time counter are combined to form an overall time base.

[0014] Furthermore, the acquisition of the measurement data in step S300 includes:

[0015] S301: Start triggering the shutter to collect data when receiving the input external pulse time signal. When the rising edge of the image sensor trigger signal is generated, the count value of the second-level time counter at this moment is the nanosecond timestamp T1 of the triggering moment;

[0016] S302: repeat step S302, the timestamp corresponding to the signal at the middle of exposure in the working state of the image sensor is T2, and the timestamp corresponding to the moment of exposure completion is T3;

[0017] S303: After the structured light sensor completes the three-dimensional data measurement, one of the exposure start time T1, the exposure middle time T2 and the exposure completion time T3 is selected and combined with the measurement data to form measurement data that records the working time information of the high-precision sensor.

[0018] Furthermore, the calibration in step S300 includes:

[0019] S304: Establishing an indicator constrained by a virtual point, after the calibration object is illuminated by the laser, extracting the calibration object contour line through the laser center line to obtain a calibration line;

[0020] S305: Determine the object-side calibration parameters according to the calibration line, and extract the virtual point positions of the image side; obtain the characteristic calibration points of the object side and the image side;

[0021] S306: Based on the feature calibration points of the object space and the image space, a corresponding relationship between the image space coordinates and the object space coordinates is established according to the coordinates of the virtual point in the phase plane and the corresponding object space coordinates.

[0022] Furthermore, the indicator has a three-dimensional structure, and its main feature is repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

[0023] Furthermore, the feature calibration points are virtualized to obtain virtual points, and the accuracy of the virtual points is ensured by the following steps:

[0024] S3041: The trapezoidal groove surface processing of the trapezoidal block needs to maintain flatness;

[0025] S3042: The inclined surfaces on both sides of the trapezoidal groove need to maintain consistent parallelism and inclination angles to ensure that the position of the virtual point does not deviate;

[0026] S3043: The edge of the trapezoidal groove can be chamfered and made of aluminum alloy, and the surface is sandblasted and oxidized after processing.

[0027] Furthermore, the virtual points of the image plane are extracted by the following steps:

[0028] S3044: Acquire an image of the original line laser irradiated on the calibration object;

[0029] S3045: applying a laser centerline extraction algorithm to obtain high-precision image coordinates of the profile of the calibration object wheel;

[0030] S3046: performing distortion correction on the obtained contour line of the calibration object to obtain a contour line without distortion;

[0031] S3047: Based on the contour line, extract the characteristic virtual points of each trapezoidal block.

[0032] Furthermore, the corresponding relationship between the image space coordinates and the object space coordinates in step S306 is:

[0033] (1)

[0034] in, is the coordinate of the virtual point in the phase plane; is the object coordinate of the virtual point.

[0035] According to a second aspect of the present invention, there is provided a timing synchronized high-frequency line structure light dynamic measurement system, comprising:

[0036] A line structured light measurement module, which is used to emit a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collect a laser line image after being reflected by the surface of the object, and extract a laser center line of the laser line image to obtain a laser center line position;

[0037] A timing module for receiving external second pulse time signals;

[0038] A synchronous core processing module, which includes an external GNSS time synchronization unit for performing second clock correction on the external second pulse time signal received by the timing module, an internal time subdivision unit for simultaneously performing high-precision time subdivision on the corresponding second clock to unify the time base of each module, and a multi-element synchronization unit for realizing time synchronization for multi-sensor synchronous control and recording the working time of each sensor;

[0039] A calibration module, used to calibrate the position of the laser center line, convert the image space coordinates into object space coordinates through a virtual point constrained calibration object and an extraction method, so as to obtain measurement data of the line structured light sensor;

[0040] and a data sorting module, which is used to package the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information.

[0041] Furthermore, the line structured light measurement module comprises:

[0042] A line laser for emitting a line laser onto an object to be measured;

[0043] and a detector for detecting the line laser image reflected by the object to be measured.

[0044] Furthermore, the synchronous core processing module includes an FPGA processing core for calibrating the PPS pulse, controlling the synchronous timing of the structured light sensor, and collecting the corresponding timestamp.

[0045] Furthermore, the calibration module includes a sensor calibration unit for calibrating the conversion relationship between image space and object space according to the line laser image.

[0046] Furthermore, the calibration module has a three-dimensional structure, and its main feature is repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

[0047] Further, the calibration module includes a plurality of virtualized characteristic calibration points, and the accuracy of the virtualized characteristic calibration points is determined by the upper and lower surfaces of two adjacent trapezoidal blocks;

[0048] The calibration module also includes a virtual point of the image side. In the extracted trapezoidal contour line, the focus of the upper bottom and the left and right waists of the trapezoidal contour, and the focus close to the center position of the laser emission are used as the virtual point of the image side.

[0049] Furthermore, the timing module includes a PPS counter synchronized with a UTC time source and a crystal counter for nanosecond counting, wherein the PPS counter is used to correct the second clock of the structured light measurement sensor at the working time, and the crystal counter is used to perform high-precision time subdivision on the corresponding second clock to unify the time reference and accurately record the working time of the structured light measurement sensor.

[0050] According to a third aspect of the present invention, a calibration platform is provided, for implementing the calibration, comprising:

[0051] A servo motor, a lead screw and a motion platform, wherein the lead screw is used to convert the rotational motion of the servo motor into the linear motion of the motion platform;

[0052] A line laser and a detector are arranged on the motion platform, the line laser irradiates along the linear motion direction, and the imaging plane of the line laser image in the camera is parallel to the image plane and parallel to the x-axis or y-axis of the image plane;

[0053] The calibration object is placed flat on the motion platform, and the line laser and the camera are arranged in a direct-irradiation manner so that the laser surface is vertically irradiated on the calibration object and is tangent to the upper surface of the trapezoidal block.

[0054] The line structured light sensor composed of the line laser and the camera moves along the motion platform under control, so that multiple frames of calibration images can be collected, and then a calibration virtual point extraction algorithm is applied to obtain a series of high-precision virtual points distributed along the laser surface. The virtual points constitute a precise calibration field.

[0055] Furthermore, the motion platform drives the line structured light sensor to continuously move along the laser surface irradiation direction, collects a series of contour lines of the calibration object under the line laser irradiation, and then extracts the high-precision virtual point position from the contour lines of these trapezoidal blocks. , combining the known dimensional design parameters of the calibration object with the moving distance of the motion platform to obtain the coordinates of the virtual point in the object space , and then the corresponding relationship between the image coordinates and the object coordinates is obtained as follows:

[0056] (1)

[0057] in, is the coordinate of the virtual point in the phase plane; is the object coordinate of the virtual point.

[0058] In general, the above technical solution conceived by the present invention has the following technical advantages compared with the prior art:

[0059] 1. The measurement method of the present invention establishes a high-precision time reference based on the sensor through a synchronous core processing method to accurately record the working time of the sensor, realizes the time synchronization of multiple sensors, and designs a special form of calibration object so that it has calibration features that can be quickly and accurately extracted. The image space coordinates are converted into object space coordinates through the virtual point constrained calibration object and extraction method, thereby obtaining the measurement data of the line structured light sensor, realizing high-frequency and high-precision measurement in a moving state.

[0060] 2. The measurement method of the present invention utilizes external GNSS signals and combines PPS pulses with a high-stability crystal oscillator to not only maintain the operation of the time base, but also accurately record the working time of the sensor with the crystal oscillator frequency as the resolution, further reducing the error of the sensor measurement time record.

[0061] 3. The measurement method of the present invention installs the calibration object on the displacement platform, moves it continuously at intervals along a fixed direction, collects high-density calibration raw data, and extracts calibration features (calibration points) in the calibration data, so that the image-space-object-space conversion relationship of the structured light sensor can be accurately established, thereby obtaining high-precision structured light sensor measurement data.

[0062] 4. The measurement method of the present invention packages the measurement data of the line structured light module with the recorded high-precision working time information, and outputs three-dimensional measurement data with accurate timestamps, which can solve the high-frequency and accurate measurement of structured light measurement sensors in dynamic environments and motion scenes, as well as the problem of coordinated operation of multiple sensors. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a schematic diagram of the existing sensor time synchronization method;

[0064] Figure 2 This is a schematic diagram of the principle of a timing synchronization high-frequency linear structured light measurement sensor device according to an embodiment of the present invention;

[0065] Figure 3This is a schematic diagram of the structure of a timing synchronization high-frequency linear structured light measurement sensor device according to an embodiment of the present invention;

[0066] Figure 4 A schematic diagram of the three-dimensional structure of the calibration object in an embodiment of the present invention;

[0067] Figure 5 A displacement calibration device in an embodiment of the present invention;

[0068] Figure 6 This is a schematic diagram of the calibration platform structure in an embodiment of the present invention;

[0069] Figure 7 A schematic flow chart of a timing synchronization high-frequency line structured light measurement method according to an embodiment of the present invention;

[0070] Figure 8 A schematic diagram of a time synchronization process in an embodiment of the present invention;

[0071] Fig. 9 This is a schematic diagram of the measurement data flow in an embodiment of the present invention;

[0072] Fig.10 It is a schematic diagram of the calibration process in an embodiment of the present invention;

[0073] Fig.11 A schematic diagram of a process for determining the accuracy of a virtual point in an embodiment of the present invention;

[0074] Fig.12 Schematic diagram of a method for determining an image-side virtual point in an embodiment of the present invention.

[0075] In all the drawings, the same figure marks represent the same technical features, specifically: 1-line laser, 2-detector, 3-data acquisition and processing unit, 4-synchronous core processing module, 5-line laser image, 6-external GNSS, 7-measurement data, 8-working time, 9-sensor calibration unit, 10-line structured light three-dimensional measurement data, 11-motion platform, 12-calibration object, 13-camera. DETAILED DESCRIPTION

[0076] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0077] Example 1

[0078] like Figure 2As shown, a timing-synchronized high-frequency line structured light measurement sensor device provided in an embodiment of the present invention is intended to improve the data collection accuracy and efficiency of line structured light sensors in applications such as high-precision three-dimensional measurement and in dynamic environments. It includes a line structured light measurement module, which is used to emit a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collect a laser line image after being reflected by the surface of the object, and extract the laser center line of the laser line image to obtain the laser center line position; a timing module for receiving an external second pulse time signal; a synchronization core processing module 4, which includes an external GNSS time synchronization unit for performing second clock correction on the external second pulse time signal received by the timing module, an internal time subdivision unit for simultaneously performing high-precision time subdivision on the corresponding second clock to unify the time base of each module, and a multi-element synchronization unit for realizing time synchronization for multi-sensor synchronous control and recording the working time of each sensor; a calibration module, which is used to calibrate the laser center line position, convert the image coordinates into object coordinates through a virtual point constrained calibration object and an extraction method, thereby obtaining the measurement data of the line structured light sensor; and a data sorting module, which is used to package the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information.

[0079] like Figure 3 As shown, further, the timing module is used to receive an external (GNSS) second pulse time signal, and provide a time reference for the entire system through precise PPS pulses and UTC time. And record the working time information for the sensor. The line structured light measurement module includes a line laser 1 for emitting a line laser to the object to be measured, emitting a beam of line laser to the surface of the object, and the laser power can be automatically adjusted according to the measured object and environment to adapt to different measurement requirements, especially suitable for high dynamic range measurement. And a detector 2 for detecting the line laser image reflected by the object to be measured, the detection range of the detector 2 is adjustable, and it cooperates with the layout angle of the laser to achieve an adjustable measurement range and accuracy, and collects the laser line image after reflection from the surface of the object. The synchronous core processing module 4 includes an FPGA processing core for calibrating the PPS pulse, controlling the synchronous timing of the structured light sensor, and collecting the corresponding timestamp. The calibration module includes a sensor calibration unit 9 for calibrating the image-to-object conversion relationship according to the line laser image. The detector 2 samples the line laser image irradiated onto the surface of the object, and the processing module in the data acquisition and processing unit 3 extracts the laser center line to obtain the sub-pixel laser center line position. After being processed by the sensor calibration unit 9, the image space coordinates are converted into object space coordinates, thereby obtaining the measurement data 7 of the line structured light sensor.

[0080] In addition, the data acquisition processing unit 3 has a high-speed data acquisition function: it uses large-scale programmable logic devices to obtain the image data collected by the detector 2. In the data acquisition processing unit 3, there is a parallelized hardware laser centerline extraction algorithm, which accurately extracts the laser centerline in the detector 2 through a high-speed and highly robust algorithm, thereby improving data processing efficiency and accuracy.

[0081] The synchronous core processing module 4 has a synchronous control function and integrates a high-precision crystal oscillator. The synchronous core processing module 4 has the functions of external time synchronization, and performs second clock correction through an external GNSS 6 signal to ensure the accuracy of time synchronization. Internal time segmentation: Use the internal crystal oscillator for high-precision time segmentation and unify the time base of each module. Multi-element synchronous control: Ensure the synchronization between the laser and multiple sensors, and accurately record the working time of the sensor 8. The synchronous core processing module 4 packages the sensor measurement data 7 with the working time of the sensor 8, and outputs the line structured light three-dimensional measurement data 10 with high-precision time information, which can provide an accurate data basis for multi-sensor data fusion.

[0082] Furthermore, the timing module receives an external second pulse time signal (such as GNSS) and provides an accurate PPS pulse. The GNSS receiver can provide accurate time information, mainly by capturing signals from multiple GNSS satellites in Earth orbit. Each satellite is equipped with an atomic clock that can send a signal with a timestamp. The receiver calculates the time difference from the transmission to the reception of the signal, and then determines the exact time. The pulse represents the starting moment of the time second, and the current accuracy can reach the ns level. This signal is used to calibrate the square wave pulse generated by the high-stability crystal oscillator. Through this method, the output of the crystal oscillator is corrected for long-term stability and maintains its high stability on a short time scale.

[0083] The module first needs to be initialized, and a differential crystal pulse counter (crystal counter) and a second time counter (PPS counter) are defined. In addition, the $GPRMC message output by the GNSS receiver is received and parsed, and the current coordinated universal time UTC (universal time coordinated) is obtained according to the "recommended positioning information" data frame to initialize the second time counter. The pulses of the 10MHz crystal input are counted to obtain the time count value in units of 100ns. The PPS signal is continuously monitored and monitored. Whenever the rising edge of the PPS signal is detected, it indicates the start of a new UTC second. When the rising edge of the PPS signal is detected, the PPS counter increases by 1 and the crystal counter is reset at the same time. The count values ​​of the differential crystal pulse counter and the second time counter are combined to form an overall time base. The PPS counter provides second-level time, while the crystal counter provides an additional 100ns-level time. Since the exposure time of the structured light sensor is at the us level or even the ms level, the timing system can meet the needs of accurately recording the working time of the line structured light sensor. The image sensor of the structured light measurement module can work in two modes: one is free shooting, in which the sensor works at the maximum frame rate allowed by the hardware performance within the set exposure time range; the other is trigger mode, in which the shutter is triggered to collect data only when the input pulse signal is received. In both working modes, the exposure start time T1, the exposure middle time T2, and the exposure completion time T3 can be recorded, as well as the frame number Count_N recorded by the frame counter at the corresponding time.

[0084] In order to obtain the nanosecond timestamp corresponding to the image sensor triggering moment, the rising edge of the image sensor triggering signal and the sensor triggering signal must be aligned to facilitate accurate acquisition of the nanosecond timestamp of the camera triggering moment, and when the rising edge of the image sensor triggering signal is generated, the count value of the system counter at that moment is the nanosecond timestamp T1 of the triggering moment. In the same way, the timestamp corresponding to the middle exposure moment signal in the working state of the image sensor is T2 and the timestamp corresponding to the exposure completion moment is T3. After the structured light module completes the three-dimensional data measurement, select one of the exposure start moment T1, the exposure middle moment T2, and the exposure completion moment T3 to combine with the measurement data to form measurement data that records the high-precision sensor working moment information.

[0085] Example 2

[0086] After completing the measurement data, the line structured light needs to be calibrated. Figure 4 and Figure 5As shown in the figure, in order to solve the problem of how to accurately calibrate the sensor, that is, to find the correspondence between the object and the image, a calibration object with virtual point constraints and an extraction method are designed. The calibration object has a three-dimensional structure, but its main feature is the repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

[0087] The calibration line is the contour line of the calibration plate extracted by the laser center line after the calibration plate is illuminated by the laser. Due to the existence of undesirable factors such as machining errors, rounded edges, uneven laser generation, divergence and propagation in space, inconsistent reflection effects on the surface of the calibration object, and the smooth characteristics of the laser center line extraction algorithm, even if the contour line extracted from a plane object after being irradiated by the laser line is not completely straight, there are varying degrees of jitter, so not all points on the calibration line participate in the calibration.

[0088] The design of the calibration plate is related to whether the correspondence between the object and image space can be accurately found, and can be divided into two parts: one is the determination of the object calibration parameters. The design of conventional calibration plates relies on high-precision processing, and the processing cost, efficiency and final accuracy are limited. In this method, the high-precision feature calibration points are virtualized, and the accuracy of the virtual points is jointly ensured by two adjacent planes, thereby converting the accuracy requirements for points or lines in conventional processing into plane accuracy control. Plane processing is easy to operate, and conventional equipment can maintain a high degree of flatness and parallelism, so that the final virtual point accuracy of the calibration plate is improved.

[0089] The calibration plate design complies with the following design principles to ensure the accuracy of the virtual points:

[0090] 1. The surface processing of the trapezoidal groove needs to maintain flatness;

[0091] 2. The inclined planes on both sides need to maintain consistent parallelism and inclination angles to ensure that the position of the virtual point does not deviate;

[0092] 3. The edges can be chamfered. Aluminum alloy is used as the material and the surface is sandblasted and oxidized after processing.

[0093] The second is the extraction of the virtual point position of the image plane. The virtual point of the image plane is the focus of the upper base and the left and right waists of the trapezoidal contour in the extracted center contour line. Usually two intersection points can be extracted from a trapezoidal block, but considering the position of the physical trapezoidal block and the irradiation angle of the laser, in each trapezoidal contour line, only the virtual intersection point closer to the center of the laser emission is used, which can reduce the adverse effects of the irradiation angle and the reflection of the object surface.

[0094] The extraction process of the image-side virtual point is as follows:

[0095] 1. Obtain the image of the original line laser irradiated on the calibration plate;

[0096] 2. Apply the laser centerline extraction algorithm to obtain high-precision image coordinates of the calibration plate contour line;

[0097] 3. Perform distortion correction on the trapezoidal plate contour obtained in step 2 to obtain a distortion-free contour;

[0098] 4. Based on the contour line, extract the characteristic virtual points of each trapezoidal block.

[0099] First, select a relatively smooth upper base of the trapezoid for straight line fitting to obtain the upper base Upline(n), (n is the trapezoid number, from the camera perspective, from left to right are 1, 2, 3...n), and the corresponding trapezoid side waist close to the laser emission center is fitted with a straight line to obtain the side waist Left / Rightline(n), (n is the trapezoid number, from the camera perspective, from left to right are 1, 2, 3...n). The intersection of Upline(n) and Left / Rightline(n) is the virtual point VP(i, j) to which the trapezoid contour block belongs.

[0100] After obtaining the calibrated feature points of the object and image space, we can calculate the coordinates of the virtual point in the image plane. And its corresponding object coordinates , and find the corresponding relationship between them.

[0101] Example 3

[0102] like Figure 6 As shown, in another embodiment of the present invention, it is proposed to construct a precision calibration platform, which is composed of a servo motor and a lead screw, and can convert the rotational motion of the servo motor into the linear motion of the translation stage, and its accuracy can reach 0.01mm. The laser and the camera are assumed to be on the translation stage, and the line laser is vertically directed downward to illuminate the calibration target placed flat on the bottom surface of the precision motion platform. Before calibration, it is necessary to ensure that there is a certain positional relationship between the laser line, the camera, and the calibration object:

[0103] 1. The line laser 1 irradiates along the moving direction of the translation stage, and the imaging plane of the line laser image 5 in the camera is parallel to the image plane and parallel to the x-axis or y-axis of the image plane.

[0104] 2. The calibration object 3 is placed flat on the base of the motion platform. The laser and camera are designed in a direct-illuminated layout so that the laser surface is vertically irradiated on the calibration object and is tangent to the upper surface of the trapezoidal block.

[0105] 3. The line structured light sensor composed of the laser and the camera moves up and down precisely along the servo electric control platform 4 under control, so as to collect multiple frames of calibration images, and then apply the aforementioned calibration virtual point extraction algorithm to obtain a series of high-precision virtual points distributed along the laser surface. These virtual points constitute a precise calibration field, in which each point is an accurate sampling of the structured light laser surface.

[0106] The precision displacement stage drives the sensor to continuously move along the direction of laser irradiation at fine intervals, and collects a series of contour lines of the calibration object under the online laser irradiation. Then, the high-precision virtual point position is extracted from the contour lines of these trapezoidal blocks. , the coordinates of the virtual point in the object space can be obtained by relying on the known dimensional design parameters of the calibration object and the moving distance of the translation stage. , and then we can solve the conversion relationship between them:

[0107] (1)

[0108] When the world coordinate system is established based on the initial motion position of the calibration object, The surface is consistent with the laser surface. For the virtual calibration space formed by the laser surface and the calibration object The above formula can be converted to:

[0109] (2)

[0110] In short:

[0111] (3)

[0112] in:

[0113] (4)

[0114] For the convenience of subsequent solutions, the matrix H is recorded as , so we have:

[0115] (5)

[0116] in

[0117] (6)

[0118] Each pair of corresponding object points and image point , which can provide a set of relations:

[0119] (7)

[0120] Recorded as

[0121] (8)

[0122] in

[0123] (9)

[0124] Since h has nine parameters and its rank is 8, when there are n ≥ 4 sets of non-collinear corresponding points, each set provides a 2×9 matrix , can form a 2n×9 matrix ,right Perform SVD decomposition, and the unit eigenvector corresponding to the minimum eigenvalue is the solution H:

[0125] (10)

[0126] Where D is a diagonal matrix with diagonal elements arranged in descending order along the diagonal, and h is the last column of V.

[0127] In the actual calibration process, the four adjacent virtual points VP(i,j), VP(i+1,j), VP(i,j+1), and VP(i+1,j+1) along the X-axis and Y-axis can form a minimum calibration unit, so as to solve the corresponding H matrix, and then use the formula

[0128] (11)

[0129] Get the image point and object point and image point In the quadrilateral formed by these four virtual points, any image point can be solved according to formula (11) to obtain the unique corresponding object point coordinates.

[0130] In the above analysis, at least four pairs of non-collinear corresponding points are needed to solve the conversion relationship between the object space and the image space. However, a set of virtual points obtained by sampling the calibration object once has a collinear relationship, which makes it difficult to provide sufficient constraints. That is, at least two samples of the calibration object at different positions are required to calibrate the measurement equation of the line structured light sensor within the entire measuring range.

[0131] Example 4

[0132] like Figure 7-12 As shown, an embodiment of the present invention provides a timing synchronized high-frequency line structured light dynamic measurement method, comprising:

[0133] S100: emitting a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collecting a laser line image after being reflected by the surface of the object, and extracting a laser center line of the laser line image to obtain a laser center line position;

[0134] S200: Receive external second pulse time signal to calibrate the second clock, and at the same time, perform high-precision time subdivision on the corresponding second clock to unify the time base of each module, realize time synchronization through multi-sensor synchronous control, and record the working time of each sensor;

[0135] S300: Calibrate the laser center line position, and convert the image space coordinates into object space coordinates through a virtual point constrained calibration object and an extraction method, so as to obtain measurement data of the line structured light sensor;

[0136] S400: Packing the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information, providing an accurate data basis for multi-sensor data fusion.

[0137] Furthermore, the time synchronization in step S200 includes:

[0138] S201: define a differential crystal pulse counter and a second-level time counter, and receive an external pulse time signal to initialize the second-level time counter;

[0139] S202: Count the external pulse time signal and continuously monitor the second time counter signal. When a rising signal of the second time counter is detected, the second time counter increases by one step and resets the differential crystal oscillator pulse counter at the same time. The count values ​​of the differential crystal oscillator pulse counter and the second time counter are combined to form an overall time base.

[0140] Furthermore, the acquisition of the measurement data in step S300 includes:

[0141] S301: Start triggering the shutter to collect data when receiving the input external pulse time signal. When the rising edge of the image sensor trigger signal is generated, the count value of the second-level time counter at this moment is the nanosecond timestamp T1 of the triggering moment;

[0142] S302: repeat step S302, the timestamp corresponding to the signal at the middle of exposure in the working state of the image sensor is T2, and the timestamp corresponding to the moment of exposure completion is T3;

[0143] S303: After the structured light sensor completes the three-dimensional data measurement, one of the exposure start time T1, the exposure middle time T2 and the exposure completion time T3 is selected and combined with the measurement data to form measurement data that records the working time information of the high-precision sensor.

[0144] Furthermore, the calibration in step S300 includes:

[0145] S304: Establishing an indicator constrained by a virtual point, after the calibration object is illuminated by the laser, extracting the calibration object contour line through the laser center line to obtain a calibration line;

[0146] S305: Determine the object-side calibration parameters according to the calibration line, and extract the virtual point positions of the image side; obtain the characteristic calibration points of the object side and the image side;

[0147] S306: Based on the feature calibration points of the object space and the image space, a corresponding relationship between the image space coordinates and the object space coordinates is established according to the coordinates of the virtual point in the phase plane and the corresponding object space coordinates.

[0148] Furthermore, the indicator has a three-dimensional structure, and its main feature is repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

[0149] Furthermore, the feature calibration points are virtualized to obtain virtual points, and the accuracy of the virtual points is ensured by the following steps:

[0150] S3041: The trapezoidal groove surface processing of the trapezoidal block needs to maintain flatness;

[0151] S3042: The inclined surfaces on both sides of the trapezoidal groove need to maintain consistent parallelism and inclination angles to ensure that the position of the virtual point does not deviate;

[0152] S3043: The edge of the trapezoidal groove can be chamfered and made of aluminum alloy, and the surface is sandblasted and oxidized after processing.

[0153] Furthermore, the virtual points of the image plane are extracted by the following steps:

[0154] S3044: Acquire an image of the original line laser irradiated on the calibration object;

[0155] S3045: applying a laser centerline extraction algorithm to obtain high-precision image coordinates of the profile of the calibration object wheel;

[0156] S3046: performing distortion correction on the obtained contour line of the calibration object to obtain a contour line without distortion;

[0157] S3047: Based on the contour line, extract the characteristic virtual points of each trapezoidal block.

[0158] Furthermore, the corresponding relationship between the image space coordinates and the object space coordinates in step S306 is:

[0159] ,

[0160] in, is the coordinate of the virtual point in the phase plane; is the object coordinate of the virtual point.

[0161] It will be easily understood by those skilled in the art that the above description is only 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 in the protection scope of the present invention.

Claims

1. A timing synchronized high-frequency line structure light dynamic measurement method, characterized in that: include: S100: emitting a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collecting a laser line image after being reflected by the surface of the object, and extracting a laser center line of the laser line image to obtain a laser center line position; S200: Receive external second pulse time signal to calibrate the second clock, and at the same time, perform high-precision time subdivision on the corresponding second clock to unify the time base of each module, realize time synchronization through sensor synchronization control, and record the working time of the sensor; S300: Calibrate the laser center line position, and convert the image space coordinates into object space coordinates through a virtual point constrained calibration object and an extraction method, so as to obtain measurement data of the line structured light sensor; S400: Packing the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information, providing an accurate data basis for multi-sensor data fusion; The time synchronization in step S200 includes: S201: define a differential crystal pulse counter and a second-level time counter, and receive an external pulse time signal to initialize the second-level time counter; S202: Count the external pulse time signal and continuously monitor the second time counter signal. When a rising signal of the second time counter is detected, the second time counter increases by one step and resets the differential crystal oscillator pulse counter at the same time. The count values ​​of the differential crystal oscillator pulse counter and the second time counter are combined to form an overall time base.

2. The timing synchronization high-frequency line structure light dynamic measurement method according to claim 1, characterized in that: The acquisition of the measurement data in step S300 includes: S301: Start triggering the shutter to collect data when receiving the input external pulse time signal. When the rising edge of the image sensor trigger signal is generated, the count value of the second-level time counter at this moment is the nanosecond timestamp T1 of the triggering moment; S302: repeat step S302, the timestamp corresponding to the signal at the middle of exposure in the working state of the image sensor is T2, and the timestamp corresponding to the moment of exposure completion is T3; S303: After the structured light sensor completes the three-dimensional data measurement, one of the exposure start time T1, the exposure middle time T2 and the exposure completion time T3 is selected and combined with the measurement data to form measurement data that records the working time information of the high-precision sensor.

3. A timing synchronized high-frequency line structured light dynamic measurement method according to claim 1 or 2, characterized in that: The calibration in step S300 includes: S304: Establishing an indicator constrained by a virtual point, after the calibration object is illuminated by the laser, extracting the calibration object contour line through the laser center line to obtain a calibration line; S305: Determine the object-side calibration parameters according to the calibration line, and extract the virtual point positions of the image side; obtain the characteristic calibration points of the object side and the image side; S306: Based on the feature calibration points of the object space and the image space, a corresponding relationship between the image space coordinates and the object space coordinates is established according to the coordinates of the virtual point in the phase plane and the corresponding object space coordinates.

4. The timing synchronization high-frequency line structure light dynamic measurement method according to claim 3 is characterized in that: The indicator has a three-dimensional structure, and its main feature is repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

5. The timing synchronization high-frequency line structure light dynamic measurement method according to claim 4, characterized in that: The feature calibration points are virtualized to obtain virtual points, and the accuracy of the virtual points is ensured by the following steps: S3041: The trapezoidal groove surface processing of the trapezoidal block needs to maintain flatness; S3042: The inclined surfaces on both sides of the trapezoidal groove need to maintain consistent parallelism and inclination angles to ensure that the position of the virtual point does not deviate; S3043: The edge of the trapezoidal groove can be chamfered and made of aluminum alloy, and the surface is sandblasted and oxidized after processing.

6. The timing-synchronized high-frequency line structure light dynamic measurement method according to claim 5, characterized in that: The virtual points of the image plane are extracted by the following steps: S3044: Acquire an image of the original line laser irradiated on the calibration object; S3045: applying a laser centerline extraction algorithm to obtain high-precision image coordinates of the profile of the calibration object wheel; S3046: performing distortion correction on the obtained contour line of the calibration object to obtain a contour line without distortion; S3047: Based on the contour line, extract the characteristic virtual points of each trapezoidal block.

7. The timing-synchronized high-frequency line structure light dynamic measurement method according to claim 6, characterized in that: The corresponding relationship between the image space coordinates and the object space coordinates in step S306 is: （1） in, is the coordinate of the virtual point in the phase plane; is the object coordinate of the virtual point.

8. A timing synchronized high-frequency line structure light dynamic measurement system, characterized in that: include: A line structured light measurement module, which is used to emit a line laser and a detection light arranged in coordination with the line laser to the surface of an object, collect a laser line image after being reflected by the surface of the object, and extract a laser center line of the laser line image to obtain a laser center line position; A timing module for receiving external second pulse time signals; A synchronous core processing module, which includes an external GNSS time synchronization unit for performing second clock correction on the external second pulse time signal received by the timing module, an internal time subdivision unit for simultaneously performing high-precision time subdivision on the corresponding second clock to unify the time base of each module, and a multi-element synchronization unit for realizing time synchronization for multi-sensor synchronous control and recording the working time of each sensor; A calibration module, used to calibrate the position of the laser center line, convert the image space coordinates into object space coordinates through a virtual point constrained calibration object and an extraction method, so as to obtain measurement data of the line structured light sensor; and a data sorting module, for packaging the measurement data with the working time of the sensor to obtain line structured light three-dimensional measurement data with high-precision time information; The time synchronization includes: Define a differential crystal pulse counter and a second time counter, receive an external pulse time signal, initialize the second time counter, count the external pulse time signal, and continuously monitor the second time counter signal. When a rising signal of the second time counter is detected, the second time counter increases a step size and resets the differential crystal pulse counter at the same time. The count values ​​of the differential crystal pulse counter and the second time counter are combined to form an overall time base.

9. The timing synchronization high-frequency line structure light dynamic measurement system according to claim 8, characterized in that: The line structured light measurement module comprises: A line laser for emitting a line laser onto an object to be measured; and a detector for detecting the line laser image reflected by the object to be measured.

10. The timing-synchronized high-frequency line-structured light dynamic measurement system according to claim 8, characterized in that: The synchronous core processing module includes an FPGA processing core for calibrating the PPS pulse, controlling the synchronous timing of the structured light sensor, and collecting the corresponding timestamp.

11. The timing synchronization high-frequency line structure light dynamic measurement system according to claim 9, characterized in that: The calibration module includes a sensor calibration unit for calibrating the conversion relationship between image space and object space according to the line laser image.

12. The timing synchronization high-frequency line structure light dynamic measurement system according to claim 10, characterized in that: The calibration module has a three-dimensional structure, and its main feature is repeated trapezoidal blocks distributed along the laser surface. The upper bottom surface of the trapezoidal block and the left and right side surfaces each form an angle surface, which intersects with the laser surface to form a laser line calibration reference surface with a continuous trapezoidal profile.

13. The timing-synchronized high-frequency line-structured light dynamic measurement system according to claim 12, characterized in that: The calibration module includes a plurality of virtualized characteristic calibration points, the accuracy of which is determined by the upper and lower surfaces of two adjacent trapezoidal blocks; The calibration module also includes a virtual point of the image side. In the extracted trapezoidal contour line, the focus of the upper bottom and the left and right waists of the trapezoidal contour, and the focus close to the center position of the laser emission are used as the virtual point of the image side.

14. A timing-synchronized high-frequency line-structured light dynamic measurement system according to any one of claims 8 to 13, characterized in that: The timing module includes a PPS counter synchronized with the UTC time source and a crystal counter for nanosecond counting. The PPS counter is used to correct the second clock of the structured light measurement sensor at the working time, and the crystal counter is used to perform high-precision time subdivision on the corresponding second clock to unify the time reference and accurately record the working time of the structured light measurement sensor.

15. A calibration platform, characterized in that: Used to implement the calibration as claimed in any one of claims 1 to 7, comprising: A servo motor, a lead screw and a motion platform, wherein the lead screw is used to convert the rotational motion of the servo motor into the linear motion of the motion platform; A line laser and a detector are arranged on the motion platform, the line laser irradiates along the linear motion direction, and the imaging plane of the line laser image in the camera is parallel to the image plane and parallel to the x-axis or y-axis of the image plane; The calibration object is placed flat on the motion platform, and the line laser and the camera are arranged in a direct-irradiation manner so that the laser surface is vertically irradiated on the calibration object and is tangent to the upper surface of the trapezoidal block; The line structured light sensor composed of the line laser and the camera moves along the motion platform under control, so that multiple frames of calibration images can be collected, and then a calibration virtual point extraction algorithm is applied to obtain a series of high-precision virtual points distributed along the laser surface. The virtual points constitute a precise calibration field.

16. A calibration platform according to claim 15, characterized in that: The motion platform drives the linear structured light sensor to continuously move along the laser surface irradiation direction, collects a series of contour lines of the calibration object under the linear laser irradiation, and then extracts the high-precision virtual point position from the contour lines of these trapezoidal blocks. , combining the known dimensional design parameters of the calibration object with the moving distance of the motion platform to obtain the coordinates of the virtual point in the object space , and then the corresponding relationship between the image coordinates and the object coordinates is obtained as follows: （1） in, is the coordinate of the virtual point in the phase plane; is the object coordinate of the virtual point.

Images