A centering method and system for optical component detection

Abstract

The present application relates to the field of optical element detection technology, and in particular to a centering method and system for optical element detection, including: acquiring a first image and a second image of an optical element from different positions respectively; performing three-dimensional reconstruction based on the first image and the second image to obtain three-dimensional point cloud data; determining the feature credibility of each point of the three-dimensional point cloud data; determining the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data, and determining the center point of the three-dimensional point cloud data based on the central axis direction; dividing the three-dimensional point cloud data into multiple groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the multiple groups of related points; and performing laser centering based on the surface point cloud data. The centering method and system for optical element detection of the present application can effectively improve the accuracy and adjustment efficiency of laser centering.

Description

Technical Field

[0001] The present invention relates to the field of optical component detection, and in particular to a centering method and system for optical component detection. Background Art

[0002] In the development of modern optical technology, the performance and precision of optical components are crucial to the effectiveness of the entire optical system. The centering of optical components refers to adjusting the position of the component to ensure that its optical axis is aligned with other components or reference axes in the system. Centering is a key step in the assembly process of optical systems because it directly affects the imaging quality and performance of the system.

[0003] At present, the commonly used method of centering optical components is to use a laser beam as a reference, detect the reflection of the laser beam on the optical component or measure the position of the optical component through an optical sensor, and then make fine adjustments. This method can achieve very high centering accuracy, but the relative adjustment process requires repeated adjustments and iterations to achieve the predetermined alignment accuracy, and the influence of environmental factors will increase the time and accuracy of the corresponding adjustment. The laser beam alignment standard in high-precision industry requires more time to adjust. Summary of the invention

[0004] In view of the above problems, the present application provides a centering method and system for optical component detection, which can effectively improve the accuracy and adjustment efficiency of laser centering.

[0005] In a first aspect, an embodiment of the present application provides a centering method for optical component detection, comprising:

[0006] Acquire a first image and a second image of the optical element from different positions respectively;

[0007] Performing three-dimensional reconstruction based on the first image and the second image to obtain three-dimensional point cloud data;

[0008] Determining the feature credibility of each point of the three-dimensional point cloud data;

[0009] Determine the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data, and determine the center point of the three-dimensional point cloud data based on the central axis direction;

[0010] Dividing the three-dimensional point cloud data into a plurality of groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the plurality of groups of related points and the feature credibility of the related points;

[0011] Performing laser centering based on the surface point cloud data;

[0012] The step of acquiring the first image and the second image of the optical element from different positions respectively comprises:

[0013] Two image acquisition devices are arranged in parallel, and a line laser is arranged at the midpoint between the two image acquisition devices;

[0014] Acquire a first image and a second image of the optical element based on line structured light scanning measurement and stereo vision measurement;

[0015] Determining the feature credibility of each point of the three-dimensional point cloud data includes:

[0016] Performing grayscale processing on the first image to obtain a third image, and performing grayscale processing on the second image to obtain a fourth image;

[0017] Matching each point of the three-dimensional point cloud data with regions of the third image and the fourth image respectively;

[0018] For each of the points, determining feature credibility based on the corresponding regions of the third image and the fourth image;

[0019] The determining feature credibility based on the corresponding regions of the third image and the fourth image comprises:

[0020] Determine the grayscale difference degree based on the grayscale mean and grayscale difference of the region in the third image and the grayscale mean and grayscale difference of the fourth image;

[0021] determining a grayscale difference reduction coefficient based on the position coordinates of the region;

[0022] Determining feature credibility based on the grayscale difference degree and the grayscale difference reduction coefficient;

[0023] The calculation formula of the grayscale difference degree is as follows:

[0024] ;

[0025] in, Indicates The grayscale difference of the area corresponding to each point is Indicates The grayscale mean of the area corresponding to the points in the third image, Indicates The grayscale mean of the area corresponding to the points in the fourth image, Indicates The grayscale variance of the area corresponding to the points in the third image is, Indicates The grayscale variance of the area corresponding to the points in the fourth image;

[0026] The calculation formula of the grayscale difference reduction coefficient is as follows:

[0027] ;

[0028] in, Indicates The grayscale difference reduction coefficient of the area corresponding to the point is Indicates The location coordinates of the area corresponding to the point, represents the position coordinates of the first image acquisition device, represents the position coordinates of the second image acquisition device, Representing coordinates and coordinates The distance Express The absolute value of is normalized.

[0029] In some embodiments, determining the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data includes:

[0030] Calculate the covariance matrix of the centralized data set based on the three-dimensional point cloud data;

[0031] Performing eigendecomposition on the covariance matrix to obtain multiple eigenvalues ​​and their corresponding eigenvectors;

[0032] The eigenvector corresponding to the largest eigenvalue is selected as the central axis direction of the three-dimensional point cloud data.

[0033] In some embodiments, the step of calculating the covariance matrix of the centralized data set based on the three-dimensional point cloud data includes:

[0034] Constructing a point cloud set data set based on the three-dimensional point cloud data;

[0035] Determine a centralized data set based on the point cloud set data set;

[0036] Determine the covariance matrix of the centered data set based on the centered data set.

[0037] In some embodiments, determining the center point of the three-dimensional point cloud data based on the central axis direction includes:

[0038] Determine the curvature difference of each sequence of the three-dimensional point cloud data in the direction of the central axis, and determine the central axis sequence based on the curvature difference;

[0039] The center point of the three-dimensional point cloud data is determined based on the central axis sequence.

[0040] In some embodiments, dividing the three-dimensional point cloud data into multiple groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the multiple groups of related points and the feature credibility of the related points includes:

[0041] Traverse the four points that are symmetrical along the central axis and the vertical line of the midpoint to obtain multiple groups of related points with the same curvature;

[0042] For each group of related points, determining a target curvature mean based on the feature credibility of the related points;

[0043] Determining a curvature adjustment value of each relevant point in each group of relevant points based on the target curvature mean value;

[0044] The curvature of each relevant point of the three-dimensional point cloud data is adjusted based on the curvature adjustment value of each relevant point to obtain the surface point cloud data of the optical element.

[0045] In some embodiments, the laser centering based on the surface point cloud data includes:

[0046] For each point of the surface point cloud data, determining estimated laser detection plane position information;

[0047] The center of the optical element is adjusted based on the estimated laser detection plane position information.

[0048] In a second aspect, an embodiment of the present application further provides a centering system for optical element detection, which includes a memory and a processor, wherein:

[0049] The memory is used to store the computer program;

[0050] The processor is used to execute the computer program to implement the centering method for optical element detection in any of the aforementioned embodiments.

[0051] The beneficial effects of the present application are as follows: the centering method and system for optical element detection of the present application adopts a projection detection method to obtain the surface configuration of the optical element, and then uses a three-dimensional reconstruction method to obtain the three-dimensional point cloud data of the surface of the optical element. Based on the three-dimensional point cloud data, laser centering is adopted to accurately obtain the position information of the laser hitting the optical element by observing the offset of the laser on the acquisition plane, and then determine the direction and distance of adjustment according to these position information and the position information of the center point. In the optimization process, the grayscale difference of the image information collected at two installation angles is used to obtain the feature credibility of the three-dimensional point cloud data point, and the corresponding central axis is determined by the symmetry characteristics of the optical element, and the position information of the point cloud data is accurately adjusted according to the curvature characteristics of each point. After processing, the optimized and adjusted point cloud data is obtained and the laser centering operation is performed. The centering method and system for optical element detection of the present application can effectively improve the accuracy and adjustment efficiency of laser centering. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 An embodiment of the present application provides a centering method for optical element detection;

[0053] Figure 2 An embodiment of the present application provides a centering system for optical component detection. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

[0055] The terms used in the implementation section of this application are only used to explain the specific embodiments of this application and are not intended to limit this application.

[0056] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0057] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0058] See also Figure 1 As shown, the embodiment of the present application discloses a centering method for optical component detection, comprising:

[0059] Step S11, acquiring a first image and a second image of the optical element from different positions respectively.

[0060] In some embodiments, acquiring the first image and the second image of the optical element from different positions respectively includes:

[0061] Two image acquisition devices are arranged in parallel, and a line laser is arranged at the midpoint between the two image acquisition devices;

[0062] A first image and a second image of the optical element are acquired based on line structured light scanning measurement and stereoscopic vision measurement.

[0063] In order to align the optical components more accurately and efficiently, it is necessary to detect the surface of the optical components, obtain the surface configuration of the corresponding optical components by using the projection detection method, and collect two-dimensional plane imaging from multiple angles to facilitate subsequent stereo matching and three-dimensional reconstruction. Therefore, it is necessary to select the corresponding detection equipment and the detection scheme related to the installation arrangement, and design the projection detection method based on the line structured light scanning measurement and stereo vision measurement method.

[0064] For example, two CMOS (Complementary Metal Oxide Semiconductor) cameras can be selected and placed in parallel, and a line laser can be placed between the two CMOS cameras to vertically bisect the line connecting the two CMOS cameras. The line structured light is projected vertically on the horizontal plane on the surface of the optical element to produce three bright thin lines, and the line structured light strip images are collected by the cameras on both sides.

[0065] Step S12: Perform three-dimensional reconstruction based on the first image and the second image to obtain three-dimensional point cloud data.

[0066] For example, stereo vision technology can be used to perform 3D reconstruction using images from two perspectives to obtain 3D point cloud data, which mainly includes the steps of feature extraction, feature matching, stereo correction, depth information recovery, point cloud generation and post-processing, and can be implemented using software or algorithms such as OpenCV and PCL (Point Cloud Library).

[0067] In order to accurately locate the center position of the optical element, the projection detection method is used to reconstruct the three-dimensional surface image of the optical element, and the two-dimensional plane imaging is collected and processed at multiple angles. However, in the specific implementation process, in order to enhance the accuracy of the final three-dimensional point cloud and accurately find the center position of the corresponding optical element, it is necessary to reduce the light pollution that occurs during the imaging process. When collecting images on both sides, if there is a certain degree of optical interference, there will be abnormal points with a certain degree of difference in the final three-dimensional point cloud set. For this, it is necessary to analyze and process the corresponding abnormal data points from the collected grayscale image and the three-dimensional point cloud set. Finally, according to the processed results, the center point position of the optical element is found and calibrated.

[0068] Step S13, determining the feature credibility of each point of the three-dimensional point cloud data.

[0069] In the above step S13, by determining the feature credibility of each point of the three-dimensional point cloud data, the influence degree of each pixel point on its optical element is obtained according to the influence of other light sources between the collected images.

[0070] In some embodiments, determining the feature credibility of each point of the three-dimensional point cloud data includes:

[0071] Performing grayscale processing on the first image to obtain a third image, and performing grayscale processing on the second image to obtain a fourth image;

[0072] Matching each point of the three-dimensional point cloud data with regions of the third image and the fourth image respectively;

[0073] For each of the points, feature credibility is determined based on the corresponding regions of the third image and the fourth image.

[0074] First, the two collected original images are gray-processed to obtain two gray-scale images.

[0075] Then, since the process of generating 3D point cloud data is to extract information from the two original images, in the corresponding 3D point cloud data, the corresponding single point corresponds to a pixel block in an area of ​​the grayscale photo, and the corresponding area size is fixed. Therefore, the points in the 3D point cloud data are matched with the areas in the two processed grayscale images.

[0076] Finally, since the corresponding image acquisition devices are not collecting images at the same position, the grayscale of the corresponding same area will be different to a certain extent. Therefore, for each point, the feature credibility is determined based on the corresponding area of ​​the third image and the fourth image.

[0077] In some embodiments, determining feature credibility based on the corresponding regions of the third image and the fourth image comprises:

[0078] Determine the grayscale difference degree based on the grayscale mean and grayscale difference of the region in the third image and the grayscale mean and grayscale difference of the fourth image;

[0079] determining a grayscale difference reduction coefficient based on the position coordinates of the region;

[0080] The feature credibility is determined based on the grayscale difference degree and the grayscale difference reduction coefficient.

[0081] When the light pollution in the corresponding grayscale area is different due to the corresponding acquisition angle, the corresponding same area has a large grayscale feature difference, which will affect the accuracy of the final three-dimensional data point. Therefore, the grayscale difference degree is determined based on the grayscale mean and grayscale difference of the area in the third image and the grayscale mean and grayscale difference of the fourth image.

[0082] For example, the grayscale situation in the region is processed, and the grayscale difference degree in the corresponding region is obtained by comparing the grayscale mean and grayscale variance of the corresponding region.

[0083] Optionally, the calculation formula for the grayscale difference degree is as follows:

[0084] ;

[0085] in, Indicates The grayscale difference of the area corresponding to each point is Indicates The grayscale mean of the area corresponding to the points in the third image, Indicates The grayscale mean of the area corresponding to the points in the fourth image, Indicates The grayscale variance of the area corresponding to the points in the third image is, Indicates The grayscale variance of the area corresponding to the points in the fourth image.

[0086] In the above calculation formula, the characteristic difference of the grayscale regions in the two corresponding images is evaluated by calculating the Euclidean distance between the grayscale variance and the grayscale mean in the same region in the two images.

[0087] The grayscale feature differences of the analyzed regions are found to be different due to the different acquisition positions of the corresponding cameras, which may cause a certain degree of change in the grayscale features of the same region. For example, the corresponding grayscale difference reduction coefficient is obtained according to the difference between the corresponding same position and the acquisition position.

[0088] Optionally, the grayscale difference reduction coefficient is calculated as follows:

[0089] ;

[0090] in, Indicates The grayscale difference reduction coefficient of the area corresponding to the point is Indicates The location coordinates of the area corresponding to the point, represents the position coordinates of the first image acquisition device, represents the position coordinates of the second image acquisition device, Representing coordinates and coordinates The distance Express The absolute value of is normalized.

[0091] In the above calculation formula, the distance difference between the calculation area and the two image acquisition devices is calculated, and the corresponding distance difference is used as the grayscale difference reduction coefficient corresponding to the area.

[0092] The feature credibility of a single point can be determined based on the grayscale difference degree obtained from the grayscale image of the corresponding area and the grayscale difference reduction coefficient obtained from the corresponding position difference. When the corresponding feature credibility is higher, the data information credibility of the three-dimensional point cloud data point corresponding to the area is higher; when the corresponding feature credibility is lower, the data information credibility of the three-dimensional point cloud data point corresponding to the area is lower.

[0093] Optionally, the calculation formula of the feature credibility is as follows:

[0094] ;

[0095] in, Indicates The feature credibility of the area corresponding to the point, Indicates The grayscale difference of the area corresponding to each point is Indicates The grayscale difference reduction coefficient of the area corresponding to the point is Indicates normalization processing.

[0096] In the above calculation formula, Indicates the degree of difference of the area corresponding to the i-th point. When the corresponding degree of difference is higher, the corresponding point credibility in the corresponding 3D point cloud data is lower. When the corresponding degree of difference is lower, the corresponding point credibility in the 3D point cloud data is higher.

[0097] Step S14: determining the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data, and determining the center point of the three-dimensional point cloud data based on the central axis direction.

[0098] Given the inherent symmetry of optical components, their center points must be located on a specific central axis. In order to accurately determine this central axis, it is necessary to determine the direction of the device's central axis, then find the corresponding axis in the direction of the central axis, and obtain all lines perpendicular to this direction. Given that these lines are symmetrical and have the same curvature on both sides, by comparing the curvature difference, the line with the smallest curvature difference is selected as the central axis. The degree of abnormality of the corresponding points is determined based on the same curvature at the same distance on the left and right sides of the corresponding central axis.

[0099] Firstly, the central axis direction of the three-dimensional point cloud data is determined based on the three-dimensional point cloud data.

[0100] In some embodiments, determining the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data includes:

[0101] Calculate the covariance matrix of the centralized data set based on the three-dimensional point cloud data;

[0102] Performing eigendecomposition on the covariance matrix to obtain multiple eigenvalues ​​and their corresponding eigenvectors;

[0103] The eigenvector corresponding to the largest eigenvalue is selected as the central axis direction of the three-dimensional point cloud data.

[0104] In some embodiments, the step of calculating the covariance matrix of the centralized data set based on the three-dimensional point cloud data includes:

[0105] Constructing a point cloud set data set based on the three-dimensional point cloud data;

[0106] Determine a centralized data set based on the point cloud set data set;

[0107] Determine the covariance matrix of the centered data set based on the centered data set.

[0108] For example, first construct a point cloud collection dataset , where each point is a three-dimensional vector. Then calculate the centroid of the point cloud dataset , subtract the centroid from each point Then we get the centralized data set Finally, based on the centralized dataset Determine the covariance matrix of the centered dataset .

[0109] Optionally, the centralized dataset The calculation formula of the covariance matrix is ​​as follows:

[0110] ;

[0111] in, represents the covariance matrix of the centered dataset, Indicates the number of 3D point cloud data. is the centralized dataset, For the matrix The transpose of .

[0112] Afterwards, the covariance matrix Perform eigendecomposition and obtain eigenvalues and the corresponding eigenvector The eigenvector corresponding to the maximum eigenvalue is selected, and the eigenvector is used as the main change direction of the corresponding three-dimensional point cloud data, that is, the central axis direction.

[0113] Then, the center point of the three-dimensional point cloud data is determined based on the central axis direction.

[0114] In some embodiments, determining the center point of the three-dimensional point cloud data based on the central axis direction includes:

[0115] Determine the curvature difference of each sequence of the three-dimensional point cloud data in the direction of the central axis, and determine the central axis sequence based on the curvature difference;

[0116] The center point of the three-dimensional point cloud data is determined based on the central axis sequence.

[0117] According to the obtained central axis direction, the sequence of the corresponding three-dimensional point cloud data in this direction is analyzed. Due to the symmetry characteristics of the optical element, when the curvature difference of the vector sequences on both sides of the sequence is the smallest, the sequence is the corresponding central axis. Therefore, the curvature difference of each sequence of the three-dimensional point cloud data in the central axis direction is determined, and the central axis sequence is determined based on the curvature difference. Once the central axis sequence is obtained, the corresponding midpoint position is the center point of the three-dimensional point cloud data.

[0118] Optionally, the calculation formula of the curvature difference is as follows:

[0119] ;

[0120] in, For the The curvature difference of the sequences is For the corresponding The curvature of the sequence, For the corresponding The curvature of the sequence, Absolute value.

[0121] In the above calculation formula, the possibility of the corresponding sequence being the central axis is evaluated by calculating the curvature difference of the sequences on both sides of the corresponding sequence. When the minimum The sequence is the central axis sequence.

[0122] Step S15: dividing the three-dimensional point cloud data into multiple groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the multiple groups of related points and the feature credibility of the related points.

[0123] The curvatures of the corresponding points on both sides of the equidistant sequence of the corresponding sequence perpendicular to the central axis should be the same, so the curvatures of the points on the corresponding rectangle are obtained for analysis and adjustment.

[0124] In some embodiments, dividing the three-dimensional point cloud data into multiple groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the multiple groups of related points and the feature credibility of the related points includes:

[0125] Traverse the four points that are symmetrical along the central axis and the vertical line of the midpoint to obtain multiple groups of related points with the same curvature;

[0126] For each group of related points, determining a target curvature mean based on the feature credibility of the related points;

[0127] Determining a curvature adjustment value of each relevant point in each group of relevant points based on the target curvature mean value;

[0128] The curvature of each relevant point of the three-dimensional point cloud data is adjusted based on the curvature adjustment value of each relevant point to obtain the surface point cloud data of the optical element.

[0129] By traversing four points that are symmetrical along the central axis and symmetrical along the midpoint perpendicular line, multiple groups of related points with the same curvature are obtained, thereby dividing the three-dimensional point cloud data into multiple groups of related points based on the central point.

[0130] Optionally, the adjustment calculation formula of the target curvature mean is as follows:

[0131] ;

[0132] in, For the The mean of the target curvature of the group-related points, For the The first The feature credibility of each point, For the The first The curvature of a point.

[0133] In the above calculation formula, the required target curvature mean is obtained by combining the feature credibility of the corresponding point and the curvature of the corresponding point using a weighted average method.

[0134] Optionally, the calculation formula of the curvature adjustment value is as follows:

[0135] ;

[0136] in, For the The first The curvature adjustment value of the point, For the The mean of the target curvature of the group-related points, For the The first The curvature of a point.

[0137] In the above calculation formula, according to the corresponding obtained The curvature adjustment value of the corresponding point is obtained by subtracting the curvature of each corresponding point from the mean target curvature of the group.

[0138] The curvature of each relevant point of the three-dimensional point cloud data is adjusted based on the curvature adjustment value of each relevant point, and the adjusted point cloud data is the surface point cloud data of the optical element.

[0139] Step S16: performing laser centering based on the surface point cloud data.

[0140] The surface point cloud data of the corresponding optical element is obtained in the previous step, and the center is further adjusted using the laser centering method.

[0141] The traditional method of laser centering is to place the optical element on a stable fixture and guide the light beam emitted by the light source to pass through the optical element. By fine-tuning the position of the fixture, the light beam is made to pass through the center of the optical element as much as possible. The detector is then used to capture the image of the light beam after passing through the optical element. When the light beam fails to accurately pass through the center, a specific angle deviation will be exhibited. For such situations, the embodiments of the present application use the method of calculating the beam deviation and angle error to continuously perform adjustment operations to further reduce the deviation of the light beam until the center of the light beam completely coincides with the center of the detector.

[0142] In some embodiments, the laser centering based on the surface point cloud data includes:

[0143] For each point of the surface point cloud data, determining estimated laser detection plane position information;

[0144] The center of the optical element is adjusted based on the estimated laser detection plane position information.

[0145] For example, for each point of the surface point cloud data, the corresponding light beam path is analyzed according to the corresponding laser incident situation, and the position information of the laser passing through the surface of the optical element in the corresponding detection plane is obtained according to the corresponding light beam path.

[0146] Optionally, the calculation formula for estimating the laser detection plane position information is:

[0147] ;

[0148] in, To estimate the position information of the laser detection plane, It is the detection plane position information when the laser does not deviate. It is the angle between the laser incident on the optical element and the vertical direction of the reflected light from the plane. It is the projection position information of the laser on the optical element. is the reflection unit vector of the laser passing through the optical element, Representing coordinates and coordinates distance.

[0149] In the above calculation formula, the laser offset vector is obtained by obtaining the distance between the incident point position and the corresponding acquisition plane and the corresponding angle and vector direction; based on the corresponding laser offset vector and the corresponding projection plane position information, the estimated laser detection plane position information of the corresponding laser on the detection plane is obtained.

[0150] Finally, the center of the optical element is adjusted based on the estimated laser detection plane position information. A detection plane is installed below the optical element, the position of the relevant points projected by the laser on the detection plane is analyzed, and then the position of the laser incident point on the optical element is obtained according to the relevant point position, so as to obtain the direction and size of the adjustment.

[0151] The present application also discloses a centering system 100 for optical component detection. Figure 2 As shown, it includes a memory 20 and a processor 10, wherein:

[0152] The memory 20 is used to store computer programs;

[0153] The processor 10 is used to execute the computer program to implement the centering method for optical element detection described in any of the above embodiments.

[0154] The memory 20, as a carrier for resource storage, can be a read-only memory, a random access memory, a disk or an optical disk, etc., and the storage method can be short-term storage or permanent storage. In addition, the centering system 100 for optical component detection also includes a power supply 30, a communication interface 40, an input / output interface 50 and a communication bus 60; wherein the power supply 30 is used to provide working voltage for each hardware device on the centering system 100 for optical component detection; the communication interface 40 can create a data transmission channel between the centering system 100 for optical component detection and external devices, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input / output interface 50 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0155] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0156] The above is a detailed introduction to the centering method and system for optical component detection provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A centering method for optical component detection, characterized in that: include: Acquire a first image and a second image of the optical element from different positions respectively; Performing three-dimensional reconstruction based on the first image and the second image to obtain three-dimensional point cloud data; Determining the feature credibility of each point of the three-dimensional point cloud data; Determine the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data, and determine the center point of the three-dimensional point cloud data based on the central axis direction; Dividing the three-dimensional point cloud data into a plurality of groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the plurality of groups of related points and the feature credibility of the related points; Performing laser centering based on the surface point cloud data; The step of acquiring the first image and the second image of the optical element from different positions respectively comprises: Two image acquisition devices are arranged in parallel, and a line laser is arranged at the midpoint between the two image acquisition devices; Acquire a first image and a second image of the optical element based on line structured light scanning measurement and stereo vision measurement; Determining the feature credibility of each point of the three-dimensional point cloud data includes: Performing grayscale processing on the first image to obtain a third image, and performing grayscale processing on the second image to obtain a fourth image; Matching each point of the three-dimensional point cloud data with regions of the third image and the fourth image respectively; For each of the points, determining feature credibility based on the corresponding regions of the third image and the fourth image; The determining feature credibility based on the corresponding regions of the third image and the fourth image comprises: Determine the grayscale difference degree based on the grayscale mean and grayscale difference of the region in the third image and the grayscale mean and grayscale difference of the fourth image; determining a grayscale difference reduction coefficient based on the position coordinates of the region; Determining feature credibility based on the grayscale difference degree and the grayscale difference reduction coefficient; The calculation formula of the grayscale difference degree is as follows: ； in, Indicates The grayscale difference of the area corresponding to each point is Indicates The grayscale mean of the area corresponding to the points in the third image, Indicates The grayscale mean of the area corresponding to the points in the fourth image, Indicates The grayscale variance of the area corresponding to the points in the third image is, Indicates The grayscale variance of the area corresponding to the points in the fourth image; The calculation formula of the grayscale difference reduction coefficient is as follows: ； in, Indicates The grayscale difference reduction coefficient of the area corresponding to the point is Indicates The location coordinates of the area corresponding to the point, represents the position coordinates of the first image acquisition device, represents the position coordinates of the second image acquisition device, Representing coordinates and coordinates The distance Express The absolute value of is normalized.

2. The centering method for optical component detection according to claim 1, characterized in that: The determining the central axis direction of the three-dimensional point cloud data based on the three-dimensional point cloud data comprises: Calculate the covariance matrix of the centralized data set based on the three-dimensional point cloud data; Performing eigendecomposition on the covariance matrix to obtain multiple eigenvalues ​​and their corresponding eigenvectors; The eigenvector corresponding to the largest eigenvalue is selected as the central axis direction of the three-dimensional point cloud data.

3. The centering method for optical component detection according to claim 2, characterized in that: The step of calculating the covariance matrix of the centralized data set based on the three-dimensional point cloud data includes: Constructing a point cloud set data set based on the three-dimensional point cloud data; Determine a centralized data set based on the point cloud set data set; Determine the covariance matrix of the centered data set based on the centered data set.

4. The centering method for optical component detection according to claim 1, characterized in that: The determining the center point of the three-dimensional point cloud data based on the central axis direction includes: Determine the curvature difference of each sequence of the three-dimensional point cloud data in the direction of the central axis, and determine the central axis sequence based on the curvature difference; The center point of the three-dimensional point cloud data is determined based on the central axis sequence.

5. The centering method for optical component detection according to claim 1, characterized in that: The method of dividing the three-dimensional point cloud data into a plurality of groups of related points based on the center point, and determining the surface point cloud data of the optical element based on the plurality of groups of related points and the feature credibility of the related points, comprises: Traverse the four points that are symmetrical along the central axis and the vertical line of the midpoint to obtain multiple groups of related points with the same curvature; For each group of related points, determining a target curvature mean based on the feature credibility of the related points; Determining a curvature adjustment value of each relevant point in each group of relevant points based on the target curvature mean value; The curvature of each relevant point of the three-dimensional point cloud data is adjusted based on the curvature adjustment value of each relevant point to obtain the surface point cloud data of the optical element.

6. The centering method for optical component detection according to claim 1, characterized in that: The laser centering based on the surface point cloud data comprises: For each point of the surface point cloud data, determining estimated laser detection plane position information; The center of the optical element is adjusted based on the estimated laser detection plane position information.

7. A centering system for optical component detection, comprising a memory and a processor, wherein: The memory is used to store the computer program; The processor is used to execute the computer program to implement the centering method for optical element detection as described in any one of claims 1 to 6.

Images