Calibration method of handheld light pen measurement system

Abstract

The invention provides a calibration method of a handheld light pen measurement system, which comprises the following steps of: calculating a target coordinate, namely acquiring an image of a light pen rotating by different angles by taking a probe as a fixed point as a mark point image, and obtaining a probe center coordinate under a camera coordinate system from the mark point image and a light pen model through a conversion strategy; a target model construction step: constructing a target model based on the standard deviation according to the probe center coordinates under the plurality of groups of camera coordinate systems; a distance parameter calculation step: solving the target model through an iterative algorithm to obtain an optimal distance parameter between the mark point and the center of the probe; a mark point coordinate optimization step, wherein mark point light pen coordinates are optimized according to the distance parameters through a bundle adjustment algorithm; and a probe center calibration step: reversely deducing a probe center light pen coordinate from the distance parameter and the optimized mark point light pen coordinate, and the method has the advantages that the calibration of the light pen measurement system can be realized only through image information without depending on a high-precision measurement tool.

Description

Technical Field

[0001] The present invention relates to the technical field of machine vision measurement, and more particularly to a calibration method for a handheld optical pen measurement system. Background Art

[0002] The handheld light pen measurement system is a three-dimensional coordinate measurement technology that combines portability, high precision, and flexibility. It is widely used in aerospace, automotive manufacturing, and parts inspection. The system usually consists of a light pen with an optical marker, an industrial camera, and a computer. Compared with traditional fixed measurement equipment, it has the advantages of free operation and suitability for measuring complex surfaces and hard-to-reach areas. It is particularly suitable for local precision inspection of large workpieces. However, in actual applications, the measurement accuracy of the light pen system is affected by many factors. For example, the insufficient machining and assembly accuracy of the light pen model, the frequent replacement of the ruby ​​probe when measuring industrial parts at different depths, and the mechanical wear of the light pen during long-term use may cause the coordinates of the marker point and the probe center to shift, thereby significantly affecting the accuracy of the solution of the probe center coordinates and reducing the measurement accuracy of the system.

[0003] Currently, calibration methods for light pen measurement systems fall into two main categories. One relies on high-precision measurement tools, such as coordinate measuring machines (CMMs), to achieve calibration by directly measuring the three-dimensional coordinates of the light pen markers. While this method can provide high measurement accuracy, it is highly dependent on external measurement equipment, has complex operational procedures, and requires high environmental requirements, limiting the flexibility of the measurement system in practical industrial applications. The other type of calibration method attempts to be independent of external measurement equipment, but these methods often have strict implementation conditions and ultimately fail to meet actual measurement accuracy requirements. Furthermore, because light pen markers often use micro-LEDs in transparent or translucent packages, the internal light-emitting chips are tiny and have certain reflective properties on the surface, making it difficult for CMMs to accurately locate their light emission centers, thereby introducing uncertainty in measurement errors. Furthermore, to accommodate the measurement needs of workpieces with varying depths and structural features, the light pen requires frequent probe replacement. The slight offsets that may be introduced during replacement further exacerbate the changes in the probe center coordinates, affecting the system's measurement accuracy. Therefore, developing a method for calibrating light pen measurement systems using image information alone, without relying on high-precision measurement tools, is of great practical significance and application value. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the purpose of the present invention is to provide a calibration method for a handheld light pen measurement system. This method establishes a mapping relationship between the light pen coordinate system and the camera coordinate system, combines multiple sets of image information to perform sensitivity analysis and optimization calculations, and obtains the optimal coordinate parameters of the marker point and the probe center, thereby achieving efficient and accurate calibration of the light pen measurement system. This not only avoids dependence on external high-precision measurement equipment, but also significantly improves the flexibility of the calibration process and measurement accuracy.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for calibrating a handheld light pen measurement system comprises the following steps: The target coordinate calculation step obtains images of the light pen rotating at different angles with the probe as a fixed point as marker point images, wherein the marker point images include the marker point and the probe center, and the marker point images and the light pen model are converted using a conversion strategy to obtain the probe center coordinates in the camera coordinate system; The target model construction step is to construct the target model based on the center coordinates of the probes in the multiple camera coordinate systems and the standard deviation; a distance parameter calculation step, solving the target model through an iterative algorithm to obtain the optimal distance parameter between the marker point and the probe center; Mark point coordinate optimization step, optimizing the marker point light pen coordinates by using the distance parameter through a bundle adjustment algorithm; In the probe center calibration step, the distance parameters and the optimized marker point light pen coordinates are used to reversely calculate the probe center light pen coordinates.

[0006] Furthermore, the target model construction step includes a sensitivity analysis strategy, which includes applying fixed perturbations to the ideal light pen coordinate components in the light pen model one by one, calculating the coordinate standard deviation of the probe center in the camera coordinate system before and after the disturbance, and then calculating the sensitivity value coefficient based on the coordinate standard deviation, and screening out the coordinate components that have a large impact on the standard deviation.

[0007] Furthermore, the sensitivity analysis strategy includes a sensitive value coefficient construction formula, and the sensitive value coefficient construction formula is configured as follows: , , , , in, is the sensitivity coefficient, are the standard deviations in the X, Y, and Z directions respectively. is the total number of captured images, To fix the perturbation size, are calculated based on the perturbed coordinates. The standard deviation of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system, , Respectively The coordinate components of the probe center in the X, Y, and Z directions after the group disturbance, , for The mean of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system.

[0008] Furthermore, the target model construction step also includes a weighted constraint strategy, which includes randomly perturbing the coordinate components that have a large impact on the standard deviation, dynamically adjusting the coordinate weights of the sensitive value coefficients through adaptive weighting, and introducing regularization constraints to construct the objective function of the target model.

[0009] Furthermore, the objective function is configured as: , , , in, is the coordinate weight, is the regularization coefficient, are 24 ideal light pen coordinate components, is the coordinate component after the second perturbation of the ideal coordinate component, is the random perturbation size.

[0010] Furthermore, the iterative algorithm includes a trust region gradient descent algorithm, and the trust region gradient descent algorithm is configured as follows: , , in, is the number of steps in the current iteration, For the next iteration, is the learning rate, The step size set for the trust region algorithm.

[0011] Furthermore, the distance parameter calculation step includes an iterative termination constraint strategy, which includes stopping the iteration when the standard deviation of the camera coordinates of each group of probe centers is less than a threshold M, outputting the disturbance variable, and calculating the distance parameters between the marker points and the marker points and the probe center based on the coordinates after the disturbance.

[0012] Furthermore, in the marker point coordinate optimization step, the distance parameter between the marker points is introduced into the bundle adjustment algorithm, and the error function is iteratively optimized by using the nonlinear least squares method to obtain the optimized marker point light pen coordinates. The error function after iterative optimization is configured as follows: , in, is the actual observed image coordinate of the marker point, is the reprojected coordinate of the marker point, is the weight factor, is the distance parameter between the landmark points, is the coordinate of each optimization, , .

[0013] Furthermore, a distance constraint formula is configured in the probe center calibration step, and the distance constraint formula is configured as follows: , in, is the probe center coordinate, is the optimized marker point light pen coordinate, is the distance parameter between the marker point and the probe center.

[0014] Furthermore, the conversion strategy includes constructing a light pen coordinate system on the light pen model, extracting the ideal light pen coordinates of the landmark point and the probe center respectively, calculating the pose matrix of the light pen coordinate system and the camera coordinate system through the PNP algorithm, and converting the ideal light pen coordinates of the probe center into the probe center coordinates in the camera coordinate system.

[0015] The beneficial effects of the present invention are as follows: by establishing a mapping relationship between the light pen coordinate system and the camera coordinate system, combining sensitivity analysis to screen key coordinates and dynamically adjust weights, and introducing regularization constraints and geometric constraints, the calibration accuracy and robustness are improved. This method does not rely on high-precision external measurement equipment and can achieve high-precision calibration of the light pen measurement system only through image information, reducing operational complexity and cost. In addition, this method is suitable for the measurement needs of workpieces with different depths and structural features, and is particularly suitable for local precision detection of complex surfaces or large workpieces, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is the overall flow chart of the present invention; Figure 2 is a schematic diagram of a light pen model in the present invention; Figure 3 It is a multi-posture collection diagram of the light pen in the present invention; Figure 4 It is a map for obtaining the coordinates of the image of the marker points in the present invention; Figure 5 It is the coordinate component diagram for the AH landmark point in the present invention; Figure 6 It is the principle diagram of bundle adjustment in the present invention; Figure 7 It is a coordinate comparison diagram before and after calibration in the present invention. DETAILED DESCRIPTION

[0017] The present invention will be described in further detail below with reference to the accompanying drawings and embodiments. Identical components are denoted by the same reference numerals. It should be noted that the terms "front," "rear," "left," "right," "upper," and "lower" used in the following description refer to directions in the accompanying drawings, and the terms "bottom," "top," "inner," and "outer" refer to directions toward or away from the geometric center of a particular component, respectively.

[0018] Since the light pen marking points are mostly made of transparent or translucent packaged micro LEDs, the internal light-emitting chip is tiny in size and has a certain reflective property on the surface, making it difficult for the three-dimensional coordinate measuring machine to accurately locate its light-emitting center, thus causing uncertainty in the measurement error. In addition, in order to meet the measurement needs of workpieces with different depths and structural features, the light pen needs to frequently replace the ruby ​​probe, and the replacement process may also introduce a slight offset, further affecting the coordinates of the probe center, thereby affecting the measurement accuracy of the system. The present invention provides a calibration method for a handheld light pen measurement system, which achieves high-precision calibration by combining image information without relying on external high-precision measurement equipment. The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, the calibration process of the present invention includes: The target coordinate calculation step requires the preparation of a conical calibration piece. Its design ensures that the spatial position of the probe center remains unchanged under different postures of the light pen. This feature provides a basic guarantee for subsequent calculations. Then, Figure 3 As shown in FIG, the light pen is placed in the conical calibration piece, and the probe is kept fixed. The light pen body is rotated to capture multiple sets of images as marker point images through an industrial camera, as shown in FIG. Figure 2 As mentioned above, the seven marker points in each marker point image are named A, B, C, D, E, F, and H respectively. After obtaining the marker point image, it needs to be preprocessed to extract the image coordinates of the marker points, as shown in the following example: Figure 4 As shown in the figure, the marker point image is first binarized, and then the image coordinates of the marker point are obtained by the ellipse fitting method of sub-pixel interpolation. This method significantly improves the accuracy of marker point center positioning by linear interpolation calculation of the grayscale values ​​of the edge pixels of the marker point. For example, in practical applications, if the grayscale distribution of the edge pixels of marker point A presents a non-uniform characteristic, its center position can be accurately determined by sub-pixel interpolation, thereby providing reliable data for subsequent calculations.

[0019] The light pen coordinate system is established with the mark point in the upper left corner of the light pen model as the origin. That is, the direction along the length of the light pen is defined as the positive direction of the X axis, the direction perpendicular to the plane of the light pen is the positive direction of the Z axis, and the Y axis direction is determined by the right-hand rule. On this basis, the ideal coordinates of the mark point AF and the probe center H in the light pen coordinate system are obtained, as shown in the following table.

[0020] According to the image coordinates of the marker points, the ideal light pen coordinates, and the camera intrinsic parameters, the PNP algorithm is used to obtain the pose information between the light pen coordinate system and the camera coordinate system. Then, combined with the ideal light pen coordinates of the probe center, the coordinates of the probe center in the camera coordinate system are obtained. That is, the rotation matrix and translation vector between the light pen coordinate system and the camera coordinate system are calculated. The calculation formula is configured as follows: , , , in, is the distortion factor used to compensate for lens distortion, are image coordinates, is the light pen coordinate, is the camera coordinate, is the intrinsic parameter matrix of the camera, is the camera coordinate of the probe center, is the light pen coordinate of the probe center, is the camera's rotation matrix, is the translation vector of the camera. The above formula can be used to accurately solve the probe center coordinates in the camera coordinate system.

[0021] The target model construction step is to construct the target model based on the standard deviation of the probe center coordinates in multiple sets of camera coordinate systems. Specifically, the target model construction step includes a sensitivity analysis strategy. The sensitivity analysis strategy includes applying a fixed perturbation to the ideal light pen coordinate components in the light pen model one by one, that is, adding a fixed perturbation with an amplitude of 0.1mm to the 24 coordinate components in sequence, perturbing only one coordinate component at a time, and keeping the other coordinates unchanged. The coordinate calculation formula after perturbation is configured as follows: ,in, is the coordinate after disturbance, are the coordinate components of 7 marker points and a probe center, , To fix the disturbance variable, we obtain the coordinates after disturbance, i.e., a new set of light pen coordinates of the marker and the probe center. We then use the PNP algorithm to calculate the pose transformation matrix from the light pen to the camera. Based on the light pen coordinates of the probe center after disturbance, we further obtain the coordinates of the probe center in the camera coordinate system. By repeating the above process for N sets of collected images, we obtain N probe center coordinates. The standard deviation is calculated in the X, Y, and Z directions respectively. The formula is configured as follows: , , , in, are the standard deviations in the X, Y, and Z directions respectively. is the total number of captured images, Respectively The coordinate components of the probe center in the X, Y, and Z directions after the group disturbance, for The mean of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system, .

[0022] The sensitivity coefficient represents the average value of the standard deviation changes in the three directions before and after the disturbance of each coordinate value. The final result is as follows Figure 5 As shown, the sensitivity coefficient construction formula is configured as follows: , in, is the sensitivity coefficient, are calculated based on the perturbed coordinates. The standard deviation of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system, , To fix the perturbation size, in actual operation, if a certain ideal light pen coordinate has a greater impact on the standard deviation of the probe center, its sensitivity coefficient is higher, indicating that this coordinate contributes more to the optimization model.

[0023] Higher weights are assigned to light pen coordinates with larger sensitivity values, and the normalized weight formula is configured as follows: , in, is the coordinate weight, The second perturbation variable for the ideal coordinate is controlled by the weight. The higher the sensitivity value, the higher the perturbation weight. The second perturbation coordinate formula is configured as: , in, is the coordinate component after the second perturbation of the ideal coordinate component, is the random perturbation size, ranging from 0.01 to 0.2.

[0024] According to the coordinates after the second ideal coordinate perturbation, that is, the new landmark and probe center coordinates, the camera coordinates of N groups of probe centers are calculated based on N groups of images, and the standard deviations in the three directions are calculated. At the same time, considering that excessive perturbations may introduce the risk of model instability or overfitting, this paper introduces a regularization term in the optimization objective function to constrain the perturbation amplitude. The final objective function formula is configured as follows: , in, is the regularization coefficient, which is used to avoid model instability or overfitting caused by excessive weight of a single parameter. The target model is composed of 24 ideal light pen coordinate components. By dynamically adjusting the weights and introducing regularization constraints, the target model can maintain high robustness under complex conditions.

[0025] In the distance parameter calculation step, the target model is solved by an iterative algorithm to obtain the optimal distance parameter between the landmark point and the probe center. Specifically, the iterative algorithm includes a trust region gradient descent algorithm, and the trust region gradient descent algorithm is configured as follows: , , in, is the number of steps in the current iteration, For the next iteration, is the learning rate, The step size set for the trust region algorithm is used to gradually approach the optimal solution, and ultimately obtain the distance parameters between the markers and between the markers and the probe center. In practical applications, the accuracy of the distance parameters directly affects the subsequent calibration results, so the convergence and stability of the iterative process must be ensured.

[0026] The distance parameter calculation step includes an iteration termination constraint strategy, which includes stopping the iteration when the standard deviation of the camera coordinates of each group of probe centers is less than the threshold M (0.05) and outputting the disturbance variable , and calculate the distance parameters between the marker points and between the marker points and the probe center based on the coordinates after disturbance.

[0027] In the step of optimizing the coordinates of the marker points, the distance parameters between the marker points are introduced into the bundle adjustment algorithm. The error function is iteratively optimized by using the nonlinear least squares method to obtain the optimized coordinates of the marker points. Figure 6 As shown, the error function after iterative optimization is configured as: , in, is the actual observed image coordinate of the marker point, is the reprojected coordinate of the marker point, is a weight factor used to balance the effects of image error and geometric constraints, is the distance parameter between the landmark points, is the coordinate of each optimization, , ,Through this algorithm, not only the light pen coordinates of the ,marking points are optimized, but also the consistency of the geometric ,relationships between the marking points is ensured, thereby improving ,the calibration accuracy.

[0028] In the probe center calibration step, the distance parameters and the optimized marker light pen coordinates are used to infer the probe center light pen coordinates. For the above-mentioned improved error function, the marker light pen coordinates are optimized using nonlinear least squares. After obtaining the optimized marker light pen coordinates, the distance parameters obtained in the first stage of calibration are combined to iteratively calculate the probe center coordinates in the calibrated light pen coordinate system. After the calibration is completed, the coordinates of the calibrated marker point and the probe center in the light pen coordinate system are output. The comparison of the coordinates before and after calibration is as follows: Figure 6 As shown, the distance constraint objective function formula is configured as: , in, is the probe center coordinate, , is the optimized marker point light pen coordinate, is the distance parameter between the marker point and the probe center, such as Figure 7 As shown in Figure 2, the comparison of coordinates before and after calibration clearly shows the calibration effect. The calibrated coordinates have higher accuracy and consistency.

[0029] Calibration principle: ①. Calibration target: 7 markers and the coordinates of the probe center in the light pen coordinate system; ②. In the first stage of calibration, the goal is to determine the distance parameter between the probe center and the mark on the light pen. Initially, the sensitivity analysis method is used to screen the ideal coordinates (the purpose of this step is to select the most sensitive coordinates and subsequently assign higher perturbation weights to accelerate the convergence of the algorithm; the method is to add a fixed perturbation of 0.1mm to the ideal coordinates to obtain a new set of markers and probe center coordinates, and then recalculate the probe center coordinates obtained from N groups of images, and calculate the standard deviation before and after the perturbation, and obtain the sensitivity coefficient according to the formula, so as to select the coordinates with high sensitivity), and then apply weighted perturbation to the selected coordinates; Then, solve the 3D coordinates of the probe center in the camera coordinate system of each set of images after the second perturbation, and calculate its standard deviation in the X, Y and Z directions. When the standard deviation in all three directions is minimized, output the perturbation variable at this time, add the perturbation variable to the ideal coordinates to obtain the perturbation coordinates, and use the distance formula between two points to calculate the distance parameters between the marker points and between the marker points and the probe center. (In fact, it is to first give the ideal coordinates a fixed perturbation to observe which points have a greater impact on the final standard deviation, and then select these points and increase the weight when perturbing for the second time. Finally, the standard deviation of the three directions of XYZ of the N sets of probe center coordinates is minimized by adjusting the size of the perturbation).

[0030] ③. Reasons for performing the second stage of calibration: The main purpose of the first stage calibration is to obtain the distance parameters between the marker point and the probe center. However, in actual applications, deviations may occur due to manufacturing and assembly errors, especially inaccurate soldering of infrared LEDs, resulting in differences between the ideal coordinates obtained by the light pen model and the actual coordinates in the real world. In addition, due to various factors such as internal and external camera parameter errors, image observation inaccuracies and position deviations of the marker points, the actual projection points on the image plane deviate from their observed positions. Therefore, in the second stage of calibration, a bundle adjustment algorithm is applied to optimize the coordinates of the marker points in the light pen coordinate system.

[0031] ④. Second stage calibration: The distance parameters between the marker points obtained in the first stage of calibration are introduced to constrain the optimization of the marker points. This step is to optimize the coordinates of the marker points by minimizing the projection error between the image coordinates of the marker points obtained by the camera and the corresponding projection points. The optimized coordinates of the marker points are combined with the distance parameters between the marker points and the probe center obtained in the first stage and iterated to obtain the optimized coordinates of the probe center. The calibration is completed.

[0032] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that improvements and modifications that do not depart from the principles of the present invention are within the scope of protection of the present invention.

Claims

1. A calibration method for a handheld light pen measurement system, characterized in that: The steps include: The target coordinate calculation step obtains images of the light pen rotating at different angles with the probe as a fixed point as marker point images, wherein the marker point images include the marker point and the probe center, and the marker point images and the light pen model are converted using a conversion strategy to obtain the probe center coordinates in the camera coordinate system; The target model construction step is to construct the target model based on the center coordinates of the probes in the multiple camera coordinate systems and the standard deviation; a distance parameter calculation step, solving the target model through an iterative algorithm to obtain the optimal distance parameter between the marker point and the probe center; Mark point coordinate optimization step, optimizing the marker point light pen coordinates by using the distance parameter through a bundle adjustment algorithm; In the probe center calibration step, the distance parameters and the optimized marker point light pen coordinates are used to reversely calculate the probe center light pen coordinates.

2. The calibration method for a handheld optical pen measurement system according to claim 1, characterized in that: The target model construction step includes a sensitivity analysis strategy, which includes applying fixed perturbations to the ideal light pen coordinate components in the light pen model one by one, calculating the coordinate standard deviation of the probe center in the camera coordinate system before and after the perturbation, and then calculating the sensitivity value coefficient based on the coordinate standard deviation, and screening out the coordinate components that have a large impact on the standard deviation.

3. The calibration method of a handheld optical pen measurement system according to claim 2, characterized in that: The sensitivity analysis strategy includes a sensitive value coefficient construction formula, and the sensitive value coefficient construction formula is configured as follows: ， ， ， ， in, is the sensitivity coefficient, are the standard deviations in the X, Y, and Z directions respectively. is the total number of captured images, To fix the perturbation size, are calculated based on the perturbed coordinates. The standard deviation of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system, , Respectively The coordinate components of the probe center in the X, Y, and Z directions after the group disturbance, , for The mean of the probe center coordinates in the X, Y, and Z directions in the group camera coordinate system.

4. The calibration method for a handheld optical pen measurement system according to claim 2, characterized in that: The target model construction step also includes a weighted constraint strategy, which includes randomly perturbing the coordinate components that have a large impact on the standard deviation, dynamically adjusting the coordinate weights of the sensitive value coefficients through adaptive weighting, and introducing regularization constraints to construct the objective function of the target model.

5. The calibration method of a handheld optical pen measurement system according to claim 4, characterized in that: The objective function is configured as: ， ， ， in, is the coordinate weight, is the regularization coefficient, are 24 ideal light pen coordinate components, is the coordinate component after the second perturbation of the ideal coordinate component, is the random perturbation size.

6. The calibration method of a handheld optical pen measurement system according to claim 5, characterized in that: The iterative algorithm includes a trust region gradient descent algorithm, and the trust region gradient descent algorithm is configured as follows: ， ， in, is the number of steps in the current iteration, For the next iteration, is the learning rate, The step size set for the trust region algorithm.

7. The calibration method of a handheld optical pen measurement system according to claim 6, characterized in that: The distance parameter calculation step includes an iterative termination constraint strategy, which includes stopping the iteration when the standard deviation of the camera coordinates of each group of probe centers is less than a threshold M, outputting the disturbance variable, and calculating the distance parameters between the marker points and the marker points and the probe center based on the coordinates after the disturbance.

8. The calibration method of a handheld optical pen measurement system according to claim 7, characterized in that: The marker point coordinate optimization step introduces the distance parameter between the marker points into the bundle adjustment algorithm, and iteratively optimizes the error function by using the nonlinear least squares method to obtain the optimized marker point light pen coordinates. The error function after iterative optimization is configured as follows: ， in, is the actual observed image coordinate of the marker point, is the reprojected coordinate of the marker point, is the weight factor, is the distance parameter between the landmark points, is the coordinate of each optimization, , .

9. The calibration method of a handheld optical pen measurement system according to claim 8, characterized in that: The probe center calibration step is configured with a distance constraint formula, which is configured as follows: ， in, is the probe center coordinate, is the optimized marker point light pen coordinate, is the distance parameter between the marker point and the probe center.

10. The calibration method of a handheld optical pen measurement system according to claim 1, characterized in that: The conversion strategy includes constructing a light pen coordinate system on the light pen model, extracting the ideal light pen coordinates of the landmark point and the probe center respectively, calculating the pose matrix of the light pen coordinate system and the camera coordinate system through the PNP algorithm, and converting the ideal light pen coordinates of the probe center into the probe center coordinates in the camera coordinate system.

Images