Handheld target visual coordinate measurement method, system, and medium based on phase features

Abstract

A handheld target visual coordinate measurement method, system and medium based on phase characteristics, the measurement method is: based on three frequency superimposed phase target feature points, three-step phase shift images are obtained by acquiring RGB three-channel images with a binocular camera, and a calibrated phase-height model is combined to realize three-dimensional reconstruction of a single area array feature circular / elliptical cone, since feature imaging is a one-to-one mapping relationship, the image feature set can be effectively avoided in the binocular camera perspective imaging process, the application also includes a system for executing the method, and a medium for storing the method; the method can improve the measurement efficiency, reduce the feature point extraction error, increase the measurement field of view, solve the influence of the imaging system defocus on the target feature point extraction, and realize the online high-precision measurement of the handheld target system.

Description

Technical Field

[0001] This invention belongs to the field of visual measurement technology, and specifically relates to a method, system and medium for measuring the visual coordinates of a handheld target based on phase features. Background Technology

[0002] With the increasing demand for surface inspection in industrial settings, portable 3D coordinate measurement technology based on vision measurement is being used more and more. Handheld target vision measurement systems typically only require one or more binocular cameras, a calibration board, a handheld target, and a dedicated probe to complete in-situ / online measurements of the surface of large and complex parts. They are less affected by the field environment, have high reliability, and have unique advantages, especially for measuring complex internal cavities and deep holes.

[0003] The handheld target is a crucial measuring device in this system. Regardless of its form, it typically consists of the following parts: visual features, a probe, a support structure connecting the visual features, and other control and adjustment auxiliary equipment. Visual features are artificially constructed auxiliary features designed to address the measurement of weak textures, invisible surfaces, and other subtle surface features. By establishing a rigid transformation relationship between the easily captured and located coordinates of the visual features and the coordinates of the probe on the target, the 3D coordinates of the probe can be determined. During measurement, the probe is manually brought into contact with the point to be measured, and a photograph is taken to achieve spatial positioning of the point. Currently, visual features are constructed primarily in two ways based on their emission patterns: passive and active. Passive features are constructed from reflective materials such as glass microspheres, forming a specific shape. The feature itself does not emit light; it relies on the reflection of an external light source. Active features are themselves a light source, such as LEDs, actively emitting light to form the feature point. Typically, active features can control the brightness of the feature point to adapt to the ambient light field of the measurement environment, thus gaining wider application.

[0004] To accurately locate the coordinates of target feature points, two conditions must be met. First, the LEDs used for "feature points" need to be small, such as surface-mount LEDs (SMD-LEDs). This results in a smaller imaging area for the target surface, facilitating coordinate extraction. Second, the feature points cannot be too small, as they become less noticeable at low resolutions, leading to larger coordinate extraction errors. Furthermore, pinhole camera imaging is a perspective imaging process, which introduces perspective distortion errors. These errors require correction based on the feature point size and the target's imaging pose to improve the accuracy of image surface coordinate extraction, making the algorithm complex.

[0005] Another key issue in the measurement process is defocusing. On the one hand, handheld target measurement systems typically measure large objects in-situ, online environments where the conditions are complex. If the handheld target image defocuses, it becomes difficult to accurately determine the image plane coordinates of feature points, leading to extraction errors that severely impact system accuracy. Therefore, a vision imaging system with a large depth of field is required. On the other hand, considering the actual conditions of industrial field measurements, imaging systems generally use industrial cameras paired with telephoto fixed-focus lenses to meet the demands of high-precision, high-resolution measurement. Considering the combined effects of focal length and aperture, to obtain a large depth of field, the aperture stop of the vision measurement system needs to be minimized. However, at this point, the image shape of the LED feature points will be significantly distorted by the modulation effect of the small aperture stop, severely affecting the image processing accuracy of extracting the center / centroid coordinates based on the shape of the LED feature point image. For example, patent application CN105678709B discloses an algorithm for correcting the optical center deviation of an LED handheld target. This algorithm analyzes the causes of optical center deviation in the imaging of LEDs as spatial feature points of the handheld target, identifies the correction direction for each feature point on the target surface, acquires a pair of target imaging planes using a binocular stereo vision system, establishes epipolar geometric constraints on the feature point pair using the fundamental matrix F, and selects the optimal correction position for the feature points through a process of coarse extraction, noise reduction, coarse positioning, and fine positioning. However, this patent application suffers from problems in both accurate feature point extraction and large depth-of-field defocus imaging, necessitating a new technology to overcome the limitations of existing accurate feature point extraction and defocus imaging on the depth of field of the measurement system. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the present invention aims to propose a handheld target visual coordinate measurement method, system, and medium based on phase features. The target feature "point" is a surface structure array with phase distribution. By separating the RGB three-channel image, a three-step phase-shifted image is obtained. Combined with the set "phase-height" model, the three-dimensional reconstruction of a circle / elliptical cone with a single surface array feature is realized. Since feature imaging is a one-to-one mapping relationship between points, it can effectively avoid extracting image feature sets during the perspective imaging process of a binocular camera. At the same time, the phase information is not sensitive to the defocus imaging of the system, reducing the influence of small aperture diaphragm shape modulation, increasing the measurement depth of field, and expanding the measurement range of the vision system.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for measuring the visual coordinates of a handheld target based on phase features includes the following steps:

[0009] Step 1: Set up a handheld target visual coordinate measurement system. This system includes a binocular camera 1 with signal connection, a handheld target 2, and a computer 3. The probe of the handheld target 2 contacts the surface to be measured, while the binocular camera 1 acquires images of the target surface features and transmits them to the computer 3. Before measurement, the binocular camera 1 is calibrated to obtain the system intrinsic parameter matrix A1, A2, ..., A n With the extrinsic parameter matrices [R1|t1], [R2|t2], ..., [R n-1 |t n-1 ], where n is the number of binocular cameras 1, and the extrinsic parameter matrix [R n-1 |t n-1 ] represents the rotation matrix R from the (n-1)th stereo camera 1 to the nth stereo camera 1. n-1 With translation matrix t n-1 The positional relationships of rigid body transformations;

[0010] Step 2: Perform probe calibration on handheld target 2 in the handheld target visual coordinate measurement system to establish the target surface feature coordinate system O. f With the probe center coordinate system O p The relative positional relationship between them is expressed by the rotation matrix R. fp Translation matrix t fp Represented as:

[0011] M p =R fp *M f +t fp

[0012] Among them, M p O is the coordinate system of the probe center. p The three-dimensional coordinates of a certain point, M f O is the coordinate system of the probe center. f The three-dimensional coordinates of a certain point;

[0013] Step 3: Use the handheld target 2 to measure the surface to be measured. During the measurement, the target probe contacts the surface at the measurement location, and at the same time, use the binocular camera 1 to acquire images of the target surface features.

[0014] Step 4: Feature extraction is performed on the acquired target surface image information. This involves demodulating the color stripe image of a single surface feature to obtain phase features. Based on the "phase-height model," circular / elliptical cones are constructed for the feature regions. An iterative algorithm is used to accurately extract the sub-pixel coordinates of the apex of the circular / elliptical cone. Based on the sub-image plane coordinates corresponding to the apex of each circular / elliptical cone in the array, a target surface feature coordinate system O is established. f The system intrinsic parameter matrix A, calibrated in step one, is used. i extrinsic parameter matrix [R] i-1 |t i-1And the rotation matrix R for probe calibration in step two. fp Translation matrix t fp Calculate the spatial coordinates of the probe's center position;

[0015] Step 5: Based on the spatial coordinates of the probe center position obtained in Step 4, and the target surface feature coordinate system O established in Step 2... f With the probe center coordinate system O p By determining the relative positional relationship between the probes, the spatial coordinates of the probe during measurement can be calculated, ultimately achieving spatial positioning of the location to be measured.

[0016] Step four specifically involves:

[0017] (4.1) For the acquired target surface image, the feature points of its color image are arranged in the following order from the inside to the outside along the radial direction:

[0018] Red, I r =A+Bcos[2πr(u,v) / T-2π / 3]

[0019] Green, I g =A + Bcos[2πr(u,v) / T]

[0020] Blue, I b =A + Bcos[2πr(u,v) / T + 2π / 3]

[0021] Color,I rgb =(I r I g I b )

[0022] Among them, I r I g I b Represents color image I rgb The phase difference between the three components of the feature point in the three channels of data is 2π / 3.

[0023] (4.2) In a color image, the green color is displayed in pixels of size P. G (u g ,v g The blue area indicates a pixel size of P. B (u b ,v b The red text indicates a pixel size of P. R (u r ,v r ), and there are:

[0024] u r =u b =2u g

[0025] v r =v b =v g =3u g

[0026] Size of a single color pixel block P RGB (u rgb ,v rgb )for:

[0027] u rgb =6u g

[0028] v rgb =v r =v b =v g =3u g

[0029] u rgb / v rgb =2:1

[0030] The period of a single feature point is set to k, where 1.5 ≤ k ≤ 3. Therefore, the display size m*n of a single feature point is:

[0031] m=P*k*u rgb

[0032] n = P * 2k * v rgb ;

[0033] Where P > 2 8 pixels; P is the number of pixels for the width of one periodic stripe.

[0034] The range of the feature point matrix parameter M*N is calculated by using the target surface area of ​​the handheld target 2 and the display size m*n of a single feature point;

[0035] (4.3) Within the M*N interval of step (4.2), the color grating feature points are captured by the binocular camera 1, and the color image I is generated using RGB three-channel information. rgb Information separation yields three color images I r I g I b ;

[0036]

[0037] The phase is as follows:

[0038]

[0039] (4.4) Based on the "phase-height" model, the phase of the feature points in the color image from step (4.3) is expanded and reconstructed into a surface shape. The reconstructed surface shape is set as a circle / elliptical cone. Therefore, the set cone height h and the physical radius R of the feature point have the following relationship:

[0040] h = R

[0041] Since the "phase-height" model is a linear model, the proportionality coefficient q is:

[0042] q=P*u rgb / 2π

[0043] Where P is the number of pixels for the width of a periodic stripe, u rgb It is a color pixel block P RGB Lateral physical dimensions;

[0044] (4.5) Unfold the reconstructed surface to obtain a circular / elliptical cone surface;

[0045] (4.6) Since there is overlap between the 2π / 3 phase difference of the color feature points, the sub-pixel center coordinates of the cone angle of the feature points are obtained by iterative calculation.

[0046] The iterative calculation in step (4.6) specifically involves:

[0047] (4.6.1) Set the iteration threshold ε;

[0048] (4.6.2) Randomly select the coordinates of at least 50% of the number of colored feature points and fit a circular / elliptical cone surface:

[0049]

[0050] Based on the fitting results, calculate the average distance ε' between the coordinates of the unsampled points and the plane passing through the central axis of the cone and the generatrix of the point.

[0051] (4.6.3) Determine the size of ε' and ε. If ε' < ε, randomly select at least 10% of the total number of colored feature points from the unselected data points, fit a circle / elliptical cone, and calculate ε'. Determine the size of ε' and ε. If 100% of the data points are selected and ε' still satisfies ε' < ε, adjust ε to ε / (2-3) and calculate from step (4.6.1).

[0052] If ε'≥ε, randomly select coordinates of points that account for no less than 50% of the number of pixels of the color feature points, and fit a circle / elliptical cone until the ε' of the sampled circle / elliptical cone is less than ε.

[0053] (4.6.4) Based on the circular / elliptical cone surface determined by the data points that finally meet the threshold requirements, extract the XOY plane sub-pixel coordinates (x, y, y) of the fitted cone apex angle.1,1 y 1,1 );

[0054] (4.6.5) Calculate the sub-pixel coordinates (x, y) of the cone apex angle of all M*N color feature points on the target surface sequentially. m,n y m,n This allows for the extraction of target surface feature points.

[0055] A handheld target visual coordinate measurement system based on phase features is installed in computer 3 and includes a processing instruction module for running each step of the above-mentioned handheld target visual coordinate method based on phase features.

[0056] A computer-readable storage medium stores a calculation program for a handheld target visual coordinate method based on phase features, wherein when the computer 3 executes the calculation program, it can implement the handheld target visual coordinate method based on phase features.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] (1) In step four of this invention, the target feature points based on the three-frequency superimposed phase are used to obtain three-step phase-shifted images by acquiring RGB three-channel images with a binocular camera. Combined with the calibrated phase-height model, the three-dimensional reconstruction of a circle / elliptical cone with a single area array feature is realized. Since feature imaging is a one-to-one mapping relationship between points, it can effectively avoid extracting image feature sets during the binocular camera perspective imaging process. At the same time, the phase information is not sensitive to the defocus imaging of the system, reducing the influence of small aperture aperture shape modulation, increasing the measurement depth of field, and expanding the measurement range of the vision system.

[0059] (2) Step (4.3) of the present invention is based on the principle of spatial phase unfolding, which has the motion state of the measured object without being restricted, and can realize real-time dynamic measurement; at the same time, it can maintain the measurement accuracy even when the large-sized object is out of focus due to the small depth of field in the fixed-focus measurement system.

[0060] In summary, this invention addresses the bottleneck in the accurate extraction of luminescent feature point coordinates during the measurement process of a handheld target measurement system. It proposes a method for the accurate extraction of area array target feature structures. This method can improve measurement efficiency, reduce feature point extraction errors, increase the measurement field of view, and solve the impact of defocusing of the imaging system on the extraction of target surface feature points, thus realizing online high-precision measurement of the handheld target system. Attached Figure Description

[0061] Figure 1 This is a schematic diagram of the handheld target visual coordinate measurement system built according to the present invention.

[0062] Figure 2 This is a schematic diagram showing the handheld target 2 and the target surface feature points.

[0063] Figure 3 This is a flowchart of the iterative algorithm in step (4.6) of the present invention. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0065] This method is a measurement approach for measuring three-dimensional objects in space using a binocular stereo vision system with a handheld target 2 constructed from characteristic phases. It aims to combine phase-shifting technology with RGB projection and three-channel filtering methods, based on the uniqueness of phase features in the transformation between photographic space and Euclidean space, to construct spatial transformation invariance of feature points on the target surface, thereby achieving rapid and accurate determination of spatial points of the object under test. Its specific implementation occurs within the measurement process of the binocular stereo vision system. The invention will be further described in detail below using a specific measurement process as an example.

[0066] A method for measuring the visual coordinates of a handheld target based on phase features includes the following steps:

[0067] Step 1: Refer to Figure 1 A handheld target visual coordinate measurement system was constructed, which includes a binocular camera 1 with signal connection, a handheld target 2, and a computer 3; the probe of the handheld target 2 contacts the surface of the target location, while the binocular camera 1 acquires images of the target surface features and transmits them to the computer 3; before measurement, the binocular camera 1 is calibrated using the BPP method proposed by Gu, FF (Calibration of stereo rigs based on the Backward Projection Process [J]. Measurement Science and Technology, 2016, 27(8): 085007), and the system intrinsic parameter matrices A1, A2, ..., A n With the extrinsic parameter matrices [R1|t1], [R2|t2], ..., [R n-1 |t n-1 ], where n is the number of binocular cameras 1, and the extrinsic parameter matrix [R n-1 |t n-1 ] represents the rotation matrix R from the (n-1)th stereo camera 1 to the nth stereo camera 1. n-1 With translation matrix t n-1 The positional relationships of rigid body transformations;

[0068] Step 2: Using the probe calibration method proposed by Ma,YY (Handheldtargetprobe tip centerposition calibration for target-based vision measurement system[J]. Measurement Science and Technology,Meas.Sci.Technol.30(2019)065013), the handheld target 2 in the handheld target vision coordinate measurement system is calibrated to establish the target surface feature coordinate system O. f With the probe center coordinate system O p The relative positional relationship between them is expressed by the rotation matrix R. fp Translation matrix t fp Represented as:

[0069] M p =R fp *M f +t fp

[0070] Among them, M p O is the coordinate system of the probe center. p The three-dimensional coordinates of a certain point, M f O is the coordinate system of the probe center. f The three-dimensional coordinates of a certain point;

[0071] Step 3: Use the handheld target 2 to measure the surface to be measured, such as deep holes, cavities, and obstructed areas. During the measurement, the target probe contacts the surface to be measured, and at the same time, use the binocular camera 1 to acquire images of the target surface features.

[0072] Step 4: Feature extraction is performed on the acquired target surface image information. This involves demodulating the color stripe image of a single surface feature to obtain phase features. Based on the "phase-height model," a circular / elliptical cone is constructed for the feature region. To eliminate the problem of inaccurate cone apex coordinates caused by crosstalk between different frequency bands, an iterative algorithm is used to accurately extract the sub-pixel coordinates of the cone apex. (Refer to...) Figure 2 Based on the sub-image plane coordinates corresponding to the apex of each circle / ellipse in the array, a target surface feature coordinate system O is established. f The system intrinsic parameter matrix A, calibrated in step one, is used. i extrinsic parameter matrix [R] i-1 |t i-1 And the rotation matrix R for probe calibration in step two. fp Translation matrix t fp The spatial coordinates of the probe's center position are calculated.

[0073] Step four specifically involves:

[0074] (4.1) For the acquired target surface image, the feature points of its color image are circular, and along the radial direction in order from the inside to the outside, they are as follows:

[0075] Red, I r =A+Bcos[2πr(u,v) / T-2π / 3]

[0076] Green, I g =A + Bcos[2πr(u,v) / T]

[0077] Blue, I b =A + Bcos[2πr(u,v) / T + 2π / 3]

[0078] Color,I rgb =(I r I g I b )

[0079] Among them, I r I g I b Represents color image I rgb The three channels of data have a phase difference of 2π / 3 between the three components of the feature point;

[0080] (4.2) In a color image, the green color is displayed in pixels of size P. G (u g ,v g The blue area indicates a pixel size of P. B (u b ,v b The red text indicates a pixel size of P. R (u r ,v r ), and there are:

[0081] u r =u b =2u g

[0082] v r =v b =v g =3u g

[0083] Size of a single color pixel block P RGB (u rgb ,v rgb )for:

[0084] u rgb =6u g

[0085] v rgb =v r =v b =v g =3u g

[0086] u rgb / v rgb =2:1

[0087] The period of a single feature point is set to k, where 1.5 ≤ k ≤ 3. Therefore, the lattice parameters of a single feature point, i.e., the physical size m*n, are:

[0088] m=P*k*u rgb

[0089] n = P * 2k * v rgb ;

[0090] To obtain a high-precision feature point phase map, the number of pixels per periodic fringe width must satisfy P > 2. 8 pixels, with the period of a single feature point set to k, 1.5≤k≤3, therefore, the point matrix parameters of a single feature point, i.e., the physical size m*n, are:

[0091] m=P*k*u rgb

[0092] n = P * 2k * v rgb

[0093] Where P > 2 8 pixels; P is the number of pixels for the width of one periodic stripe.

[0094] The range of the feature point matrix parameter M*N is calculated by using the target surface area of ​​the handheld target 2 and the display size m*n of a single feature point;

[0095] (4.3) Within the M*N interval of step (4.2), the color grating feature points are captured by the binocular camera 1, and the color image I is generated using RGB three-channel information. rgb Information separation yields three color images I r I g I b Based on step (4.1), since the phase difference of the feature point units in the three color images is 2π / 3, calculate their wrapping phase and unfolding phase:

[0096] The package phase is:

[0097]

[0098] The phase is as follows:

[0099]

[0100] (4.4) In order to obtain more accurate sub-pixel center coordinates X(x,y), a "phase-height" model is needed to reconstruct the surface shape of the unfolded phase in step (4.3). In order to maintain the consistency of the horizontal and vertical positioning accuracy of the target surface, the reconstructed surface shape is set as a circle / elliptical cone. Therefore, the cone height h and the physical size radius R of the feature point need to be set, with the following relationship:

[0101] h = R

[0102] Therefore, the "phase-height" model is a linear model, and the scaling factor q is:

[0103] q=P*u rgb / 2π

[0104] Where P is the number of pixels for the width of a periodic stripe, u rgb It is a color pixel block P RGB Lateral physical dimensions.

[0105] (4.5) Since the set surface shape is a circle / elliptical cone, the phase of any generatrix is ​​continuous and linear. Therefore, the reconstructed surface shape is linearly expanded to obtain a circle / elliptical cone surface shape.

[0106] (4.6) Considering the overlap between the 2π / 3 phase differences of color feature points, which leads to crosstalk between the color channels of the stereo camera 1, there will be reconstruction phase errors in the areas where different colors overlap, resulting in water ripples in the corresponding reconstruction surface area. In order to obtain the accurate sub-pixel center coordinates of the cone angle, the following iterative algorithm is used for calculation:

[0107] (4.6.1) Set the iteration threshold ε. Based on the resolution R of the binocular camera 1, the effective size of the handheld target 2, the number of feature points M*N, and the maximum value of phase deviation caused by crosstalk between color channels, the threshold ε is selected based on experience.

[0108] (4.6.2) Randomly select the coordinates of 50% of the number of pixels occupied by the colored feature points, and fit a circular / elliptical cone surface:

[0109]

[0110] Calculate the average distance ε' between the coordinates of the unsampled points and the plane passing through the central axis of the cone and the generatrix of the point;

[0111] (4.6.3) Determine the size of ε' and ε. If ε' < ε, randomly select 10% of the total number of colored feature points from the unselected data points, fit a circle / elliptical cone, and calculate ε'. Determine the size of ε' and ε. If 100% of the data points are selected and ε' still satisfies ε' < ε, adjust ε to ε / 2 and calculate from step (4.6.1).

[0112] If ε'≥ε, randomly select the coordinates of 50% of the number of pixels occupied by the color feature points, and fit a circle / elliptical cone until the ε' of the sampled circle / elliptical cone is less than ε.

[0113] (4.6.4) Based on the circular / elliptical cone surface determined by the data points that finally meet the threshold requirements, extract the XOY plane sub-pixel coordinates (x, y, y) of the fitted cone apex angle. 1,1 y 1,1 );

[0114] (4.6.5) Calculate the sub-pixel coordinates (x, y) of the cone apex angle of all M*N color feature points on the target surface sequentially. m,n y m,n This allows for the extraction of target surface feature points.

[0115] Step 5: Based on the spatial coordinates of the probe center position obtained in Step 4, and the target surface feature coordinate system O established in Step 2... f With the probe center coordinate system O p By determining the relative positional relationship between the probes, the spatial coordinates of the probe during measurement can be calculated, ultimately achieving spatial positioning of the location to be measured.

[0116] The present invention also includes a handheld target visual coordinate measurement system and medium based on phase features, specifically:

[0117] A handheld target visual coordinate measurement system based on phase features is installed in computer 3 and includes a processing instruction module for running each step of the above-mentioned handheld target visual coordinate method based on phase features.

[0118] A computer-readable storage medium stores a calculation program for a handheld target visual coordinate method based on phase features, wherein when the computer 3 executes the calculation program, it can implement the handheld target visual coordinate method based on phase features.

[0119] This invention does not target any particular type of display device for the handheld target 2, whether it is an LCD screen, LED, microLED, or OLED, nor does it target any measurement system, whether it is monocular, binocular, or multi-view; nor does it target any arrangement of phase feature points in an array, such as triangles, squares, rings, or other irregular shapes, as long as the array is constructed using these feature points. In other words, any use of this feature structure and measurement method in the visual measurement system for the handheld target 2 can be considered an extension and variation of this method as its core.

Claims

1. A method for measuring the visual coordinates of a handheld target based on phase features, characterized in that, Includes the following steps: Step 1: Set up a handheld target visual coordinate measurement system. The system includes a binocular camera (1) with signal connection, a handheld target (2) and a computer (3); the probe of the handheld target (2) contacts the surface of the position to be measured, and at the same time, the binocular camera 1 acquires images of the target surface features and transmits them to the computer (3); before measurement, the binocular camera (1) is calibrated to obtain the system intrinsic parameter matrix. With extrinsic matrix Where n is the number of binocular cameras (1), and the extrinsic parameter matrix is... This represents the rotation matrix from the (n-1)th binocular camera (1) to the nth binocular camera (1). With translation matrix The positional relationships of rigid body transformations; Step 2: Perform probe calibration on the handheld target (2) in the handheld target visual coordinate measurement system and establish the target surface feature coordinate system. coordinate system with probe center The relative positional relationship between them, wherein the relative positional relationship is expressed by a rotation matrix. Translation matrix Represented as: in, The coordinate system of the probe center The three-dimensional coordinates of a certain point The coordinate system of the probe center The three-dimensional coordinates of a certain point; Step 3: Use a handheld target (2) to measure the surface to be measured. During the measurement, the target probe contacts the surface to be measured, and at the same time, use a binocular camera (1) to acquire images of the target surface features. Step 4: Feature extraction is performed on the acquired target surface image information. This involves demodulating the color stripe image of a single surface feature to obtain phase features. Based on the "phase-height model," circular / elliptical cones are constructed for the feature regions. An iterative algorithm is used to accurately extract the sub-pixel coordinates of the apex of the circular / elliptical cone. Based on the sub-image plane coordinates corresponding to the apex of each circular / elliptical cone in the array, a target surface feature coordinate system is established. The system intrinsic parameter matrix calibrated in step one extrinsic parameter matrix And the rotation matrix for probe calibration in step two. Translation matrix Calculate the spatial coordinates of the probe's center position; Step 5: Based on the spatial coordinates of the probe center position obtained in Step 4, and the target surface feature coordinate system established in Step 2. coordinate system with probe center By determining the relative positional relationship between them, the spatial coordinates of the probe during measurement can be calculated, thus achieving spatial positioning of the location to be measured. Step four specifically involves: (4.1) For the acquired target surface image, the feature points of its color image are arranged in the following order from the inside to the outside along the radial direction: red, green, blue, color, in, , , Represents color image The phase difference of the three components of the feature point in the three channels of data is... ; (4.2) In a color image, the green pixel size is [missing information]. The blue area indicates a pixel size of The red display shows a pixel size of And there are: Size of a single color pixel block for: Therefore, the lattice parameters of a single feature point, i.e., the physical size m*n, are: ； in, ; The number of pixels representing the width of a single periodic stripe; The range of the feature point matrix parameter M*N is calculated by using the target surface area of ​​the handheld target (2) and the display size m*n of a single feature point; (4.3) Within the M*N interval of step (4.2), the color grating feature points are captured by the binocular camera (1), and the color image is generated by the RGB three-channel information. Information separation yielded three color images. , , Based on step (4.1), due to the phase difference of the feature point units in the three color images... Calculate its wrapping phase and unfolding phase: The package phase is: The phase is as follows: (4.4) Based on the "phase-height" model, the phase of the feature points in the color image from step (4.3) is expanded and reconstructed into a surface shape. The reconstructed surface shape is set as a circle / elliptical cone. Therefore, the set cone height h and the physical radius R of the feature point have the following relationship: Since the "phase-height" model is a linear model, the proportionality coefficient q is: Where P is the number of pixels representing the width of a periodic stripe. It is a block of colored pixels. Lateral physical dimensions; (4.5) Unfold the reconstructed surface to obtain a circular / elliptical cone surface; (4.6) Due to color feature points If there is overlap between phase differences, the sub-pixel center coordinates of the cone angle of the feature point are obtained through iterative calculation.

2. The handheld target visual coordinate measurement method based on phase features according to claim 1, characterized in that, The iterative calculation in step (4.6) includes the following steps: (4.6.1) Set the iteration threshold ; (4.6.2) Randomly select the coordinates of at least 50% of the number of colored feature points and fit a circular / elliptical cone surface: Based on the fitting results, calculate the average distance between the coordinates of the unsampled points and the plane passing through the central axis of the cone and the generatrix of the point. ; (4.6.3) Judgment Size, if Randomly select at least 10% of the total number of colored feature points from the unselected data points, fit a circular / elliptical cone surface, and calculate... ,judge Size; if 100% of the data points are extracted, it still satisfies the requirement. ,Adjustment for / (2-3), and calculate from step (4.6.1); like Randomly select coordinates of at least 50% of the pixels occupied by the colored feature points, and fit them to a circle / elliptical cone. Repeat this process until the selected samples fit a circle / elliptical cone. Less than until; (4.6.4) Based on the circular / elliptical cone surface determined by the data points that finally meet the threshold requirements, extract the XOY plane sub-pixel coordinates of the fitted cone apex angle. ; (4.6.5) Calculate the sub-pixel coordinates of the cone apex angles of all M*N color feature points on the target surface sequentially. This enables the extraction of target surface feature points.

3. A handheld target visual coordinate measurement system based on phase features, characterized in that, Installed in a computer (3), including a processing instruction module that runs each step of the handheld target visual coordinate method based on phase features as described in any one of claims 1-2.

4. A computer-readable storage medium, characterized in that, The computer (3) stores a calculation program for the handheld target visual coordinate method based on phase features as described in any one of claims 1-2, and the computer (3) can implement the handheld target visual coordinate method based on phase features when executing the calculation program.

Images