Binocular camera distance adjusting method and system for flange plate measurement

Abstract

The invention relates to the technical field of binocular camera spacing adjustment, and provides a binocular camera spacing adjustment method and system for flange plate measurement, and the method comprises the steps: obtaining a measurement task data set of a to-be-measured flange plate; based on a binocular camera, according to the measurement task data set, obtaining a binocular vision image of the to-be-measured flange plate; based on the binocular vision image, utilizing a binocular camera measurement model to obtain a depth distance and a baseline distance of the binocular camera; and based on the depth distance and the baseline distance, binocular camera spacing adjustment is realized. The binocular vision image of the to-be-measured flange plate is acquired based on the binocular camera according to the measurement task data set of the to-be-measured flange plate, the depth distance and the baseline distance of the binocular camera are acquired by analyzing the binocular vision image, and the distance between the binocular camera is automatically adjusted based on the depth distance and the baseline distance, so that manual adjustment errors can be reduced, and the measurement accuracy is improved. And the efficiency and the quality of binocular camera distance adjustment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of binocular camera spacing adjustment, and in particular to a binocular camera spacing adjustment method and system for flange measurement. Background Art

[0002] Flange connection is to fix two pipes, pipe fittings or equipment on a flange plate, add flange gaskets between the two flange plates, and fasten them together with bolts to complete the connection. Some pipe fittings and equipment already have flange plates, which also belong to flange connection. Flange connection is an important connection method for pipeline construction. Flange connection is easy to use and can withstand greater pressure. Flange connection is widely used in industrial pipelines.

[0003] A flange is a component that connects pipes and equipment. It is mainly composed of two flanges and a sealing gasket in the middle. After the flange is produced, its appearance, size, thickness and other parameters need to be measured. Among them, the thickness of the flange is related to whether the flange can withstand the required pressure and load during use, and avoid safety accidents such as rupture and leakage of the flange. Therefore, the thickness measurement of the flange is extremely important.

[0004] In the process of automobile parts production, with the development of intelligence in the automobile industry, the demand for unmanned and intelligent production workshops is increasing. Among them, the identification and positioning technology of workpieces has always been an important research content in the field of intelligent automobile manufacturing. The application of binocular cameras for identification and positioning is becoming more and more widespread. In the process of parts assembly, due to the wide variety of parts, employees will constantly check and compare in order to ensure the accuracy of assembly. Among them, the use of binocular cameras to measure flanges is a work content that needs to be carried out continuously. Since the camera is binocular, its baseline distance, object distance and shooting angle determine whether the image can be completely within the lens range, so occlusion occurs, that is, for one camera, the complete object can be photographed, while for the other camera, the complete object cannot be photographed. This requires the spacing of the binocular cameras to be adjusted to improve the efficiency and accuracy of flange measurement.

[0005] Therefore, it is necessary to provide a binocular camera spacing adjustment method and system for flange measurement. Summary of the invention

[0006] The present invention provides a binocular camera spacing adjustment method and system for flange measurement. According to a measurement task data set of a flange to be measured, a binocular visual image of the flange to be measured is obtained based on a binocular camera, and the depth distance and baseline distance of the binocular camera are obtained by analyzing the binocular visual image. Based on the depth distance and the baseline distance, automatic adjustment of the binocular camera spacing is achieved, which can reduce manual adjustment errors and improve the efficiency and quality of binocular camera spacing adjustment.

[0007] The present invention provides a binocular camera spacing adjustment method for flange measurement, comprising:

[0008] Obtaining a measurement task data set of the flange to be measured;

[0009] Based on the binocular camera and the measurement task data set, the binocular vision image of the flange to be measured is obtained;

[0010] Based on the binocular vision image, the binocular camera measurement model is used to obtain the depth distance and baseline distance of the binocular camera;

[0011] Based on the depth distance and baseline distance, the binocular camera distance adjustment is achieved.

[0012] Furthermore, a measurement task data set of the flange to be measured is obtained, including:

[0013] Obtain the model and parameter data of the flange to be measured;

[0014] According to the model and parameter data, based on the set measurement task process, a measurement task data set is generated; the measurement task data set is used to measure the flange to be measured according to the set test sequence.

[0015] Further, according to the model and parameter data, based on the set measurement task process, a measurement task data set is generated, including: according to the model and parameter data, a keyword of the flange to be measured is obtained;

[0016] Import keywords into the pre-set keyword analysis model and train to generate keyword vectors;

[0017] Based on the keyword vectors, the K-means algorithm is used to perform cluster analysis to obtain several word vector groups, and multiple groups of measurement task items are generated based on the word vector groups;

[0018] Generate a first measurement ranking according to the measurement priority of the measurement task item;

[0019] Generate a second measurement sorting according to the binocular camera spacing adjustment amplitude value corresponding to the measurement task item;

[0020] After weighted accumulation of the first measurement ranking and the second measurement ranking, a third measurement ranking is generated;

[0021] Based on the third measurement sequence, a measurement task process is set, and a measurement task data set arranged according to the third measurement sequence is generated.

[0022] Furthermore, based on the binocular camera and according to the measurement task data set, a binocular visual image of the flange to be measured is obtained, including:

[0023] Using the left camera of the binocular camera to capture a first image, preprocessing the first image to obtain a preprocessed first image, and extracting features from the preprocessed first image to obtain features of the first image;

[0024] Using the right camera of the binocular camera to capture a second image, preprocessing the second image to obtain a preprocessed second image, performing feature extraction on the preprocessed second image to obtain second image features, and extracting second feature points from the second image features;

[0025] Based on the set image feature library, the reference image feature is extracted, and based on the SIFT and Hu feature fusion algorithm, it is determined whether the first image feature matches the reference feature; if not, the left camera of the binocular camera is used to capture the first image; if they match, the first feature point in the first image feature is extracted;

[0026] Perform stereo matching on the first feature point and the second feature point to obtain matching feature points;

[0027] After the binocular camera is calibrated, three-dimensional reconstruction is performed based on the matching feature points to obtain the binocular visual image.

[0028] Further, stereo matching is performed on the first feature point and the second feature point to obtain matching feature points, including:

[0029] Selecting a target feature point from the first feature points, and selecting a sub-window on the area where the target feature point is located;

[0030] According to the set similarity determination method, a target sub-window to be determined of several target feature points to be determined in the second feature point is determined. If the target sub-window to be determined has the largest similarity, the target sub-window to be determined is used as the target selection window, and the corresponding feature points summarized in the target selection window are used as matching feature points.

[0031] Furthermore, after the binocular camera is calibrated, three-dimensional reconstruction is performed based on matching feature points to obtain a binocular visual image, including:

[0032] The binocular camera is pre-calibrated using Zhang Zhengyou calibration method to obtain the camera's internal and external parameters;

[0033] According to the matching feature points, the local point cloud information is calculated, and the three-dimensional coordinate information of the flange to be measured is restored by point cloud stitching;

[0034] Based on the three-dimensional coordinate information, a binocular vision image is constructed and generated.

[0035] Furthermore, based on the binocular vision image, the binocular camera measurement model is used to obtain the depth distance and baseline distance of the binocular camera, including:

[0036] Extract feature points from the binocular vision image to obtain a feature point set;

[0037] A binocular camera measurement model is established based on the correspondence between feature point sets, and depth information and camera posture are calculated;

[0038] According to the depth information and camera pose, the depth distance and baseline distance are obtained.

[0039] Furthermore, based on the depth distance and the baseline distance, the binocular camera spacing adjustment is implemented, including:

[0040] Based on the depth distance and the baseline distance, a distance data set is established;

[0041] Based on the distance data set, a matching relationship library between the model and parameter data of the flange to be measured and the distance data is constructed;

[0042] According to the model and parameter data of the flange to be measured, the matching distance data is obtained by using the matching correspondence library;

[0043] Adjust the binocular camera distance based on the distance data.

[0044] Furthermore, based on the depth distance and the baseline distance, the binocular camera spacing adjustment is realized, and a binocular camera spacing adjustment control model is constructed to efficiently and accurately adjust the binocular camera spacing, specifically:

[0045] Based on the distance data, according to the measurement order of the flange to be measured, the adjustment difference of each distance adjustment is obtained;

[0046] According to the adjustment difference, obtain the left camera adjustment difference and the right camera adjustment difference;

[0047] Construct a control mechanism and a processor connected to the binocular camera; the control mechanism is used to control and adjust the distance of the left camera and the distance of the right camera, and the processor is used to generate control commands and control the control mechanism;

[0048] Using the left camera adjustment difference and the right camera adjustment difference, a corresponding control command set is constructed;

[0049] Using the processor, according to the control command set, a binocular camera distance adjustment control model is constructed to control the control mechanism and adjust the binocular camera distance to meet the measurement task requirements of the flange to be measured. Among them, the binocular camera distance adjustment control model is generated based on artificial neural network training.

[0050] Binocular camera spacing adjustment system for flange measurement, including:

[0051] A measurement task data set acquisition module is used to acquire a measurement task data set of a flange to be measured;

[0052] A binocular vision image acquisition module is used to acquire a binocular vision image of the flange to be measured based on a binocular camera and a measurement task data set;

[0053] A depth distance and baseline distance acquisition module is used to obtain the depth distance and baseline distance of the binocular camera based on the binocular vision image and using the binocular camera measurement model;

[0054] The binocular camera distance adjustment implementation module is used to adjust the binocular camera distance based on the depth distance and the baseline distance.

[0055] Compared with the prior art, the present invention has the following advantages and beneficial effects: according to the measurement task data set of the flange to be measured, based on the binocular camera, the binocular visual image of the flange to be measured is obtained, and by analyzing the binocular visual image, the depth distance and baseline distance of the binocular camera are obtained, and based on the depth distance and baseline distance, the binocular camera spacing is automatically adjusted, which can reduce manual adjustment errors and improve the efficiency and quality of binocular camera spacing adjustment.

[0056] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0057] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0059] Figure 1 A schematic diagram of the steps of a binocular camera spacing adjustment method for flange measurement;

[0060] Figure 2 Schematic diagram of the method steps for obtaining the measurement task data set of the flange to be measured;

[0061] Figure 3 Schematic diagram of the structure of the binocular camera spacing adjustment system used for flange measurement. DETAILED DESCRIPTION

[0062] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0063] The present invention provides a binocular camera spacing adjustment method for flange measurement, such as Figure 1 As shown, including:

[0064] Obtaining a measurement task data set of the flange to be measured;

[0065] Based on the binocular camera and the measurement task data set, the binocular vision image of the flange to be measured is obtained;

[0066] Based on the binocular vision image, the binocular camera measurement model is used to obtain the depth distance and baseline distance of the binocular camera;

[0067] Based on the depth distance and baseline distance, the binocular camera distance adjustment is achieved.

[0068] The working principle of the above technical solution is as follows: in order to realize the binocular camera spacing adjustment for flange measurement, the present invention first obtains a measurement task data set of the flange to be measured; then, based on the binocular camera, obtains a binocular visual image of the flange to be measured according to the measurement task data set; further, based on the binocular visual image, uses the binocular camera measurement model to obtain the depth distance and baseline distance of the binocular camera; finally, based on the depth distance and baseline distance, realizes the binocular camera spacing adjustment.

[0069] The beneficial effects of the above technical solution are: by adopting the solution provided in this embodiment, based on the measurement task data set of the flange to be measured, a binocular visual image of the flange to be measured is obtained based on the binocular camera, and by analyzing the binocular visual image, the depth distance and baseline distance of the binocular camera are obtained, and based on the depth distance and baseline distance, the binocular camera spacing is automatically adjusted, which can reduce manual adjustment errors and improve the efficiency and quality of binocular camera spacing adjustment.

[0070] In one embodiment, Figure 2 As shown, the measurement task data set of the flange to be measured is obtained, including:

[0071] Obtain the model and parameter data of the flange to be measured;

[0072] According to the model and parameter data, based on the set measurement task process, a measurement task data set is generated; the measurement task data set is used to measure the flange to be measured according to the set test sequence.

[0073] The working principle of the above technical solution is: in order to obtain the measurement task data set of the flange to be measured, the present invention first obtains the model and parameter data of the flange to be measured; then, according to the model and parameter data, based on the set measurement task process, a measurement task data set is generated; the measurement task data set is used to measure the flange to be measured according to the set test sequence.

[0074] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by obtaining the measurement task data set of the flange to be measured, a basis and conditions are provided for the implementation of subsequent measurements.

[0075] In one embodiment, generating a measurement task data set based on the set measurement task process according to the model and parameter data includes: obtaining keywords of the flange to be measured according to the model and parameter data;

[0076] Import keywords into the pre-set keyword analysis model and train to generate keyword vectors;

[0077] Based on the keyword vectors, the K-means algorithm is used to perform cluster analysis to obtain several word vector groups, and multiple groups of measurement task items are generated based on the word vector groups;

[0078] Generate a first measurement ranking according to the measurement priority of the measurement task item;

[0079] Generate a second measurement sorting according to the binocular camera spacing adjustment amplitude value corresponding to the measurement task item;

[0080] After weighted accumulation of the first measurement ranking and the second measurement ranking, a third measurement ranking is generated;

[0081] Based on the third measurement sequence, a measurement task process is set, and a measurement task data set arranged according to the third measurement sequence is generated.

[0082] The working principle of the above technical solution is: in order to generate a measurement task data set based on the model and parameter data and the set measurement task process, the present invention first obtains the keywords of the flange to be measured according to the model and parameter data; then imports the keywords into a pre-set keyword analysis model, trains and generates keyword vectors; then based on the keyword vectors, a cluster analysis is performed using the K-means algorithm to obtain several word vector groups, and multiple groups of measurement task items are generated based on the word vector groups; then a first measurement ranking is generated according to the measurement priority of the measurement task item; a second measurement ranking is generated according to the size of the binocular camera spacing adjustment amplitude value corresponding to the measurement task item; a third measurement ranking is generated after weighted accumulation of the first measurement ranking and the second measurement ranking; finally, based on the third measurement ranking, a measurement task process is set, and a measurement task data set arranged according to the third measurement ranking is generated.

[0083] The beneficial effect of the above technical solution is: by adopting the solution provided by this embodiment, the measurement task data set is generated according to the model and parameter data based on the set measurement task process, so that the accuracy of the measurement task data set generation can be guaranteed.

[0084] In one embodiment, based on a binocular camera and according to a measurement task data set, a binocular visual image of a flange to be measured is obtained, including:

[0085] Using the left camera of the binocular camera to capture a first image, preprocessing the first image to obtain a preprocessed first image, and extracting features from the preprocessed first image to obtain features of the first image;

[0086] Using the right camera of the binocular camera to capture a second image, preprocessing the second image to obtain a preprocessed second image, performing feature extraction on the preprocessed second image to obtain second image features, and extracting second feature points from the second image features;

[0087] Based on the set image feature library, the reference image feature is extracted, and based on the SIFT and Hu feature fusion algorithm, it is determined whether the first image feature matches the reference feature; if not, the left camera of the binocular camera is used to capture the first image; if they match, the first feature point in the first image feature is extracted;

[0088] Perform stereo matching on the first feature point and the second feature point to obtain matching feature points;

[0089] After the binocular camera is calibrated, three-dimensional reconstruction is performed based on the matching feature points to obtain the binocular visual image.

[0090] The working principle of the above technical solution is as follows: in order to obtain a binocular visual image of the flange to be measured, the present invention first uses the left camera of the binocular camera to capture a first image, and preprocesses the first image to obtain a preprocessed first image, performs feature extraction on the preprocessed first image, and obtains first image features; then uses the right camera of the binocular camera to capture a second image, and preprocesses the second image to obtain a preprocessed second image, performs feature extraction on the preprocessed second image, obtains second image features, and extracts second feature points from the second image features; based on a set image feature library, extracts reference image features, and based on SIFT and Hu feature fusion algorithms, determines whether the first image features match the reference features; when the SIFT algorithm is used for matching, when the image is rotated When the feature points are deviated, the matching will be affected. According to the shortcomings of SIFT feature points in matching the scattered workpieces of flanges, such as many mismatching points and long time consumption, the concept of coarse matching is introduced, that is, by adding Hu moment features of multiple angles, as many similar contours as possible are found without losing accuracy in the coarse matching stage, and the area similar to the flange target contour is screened out. Then, holes are extracted in this area to reduce feature points, thereby reducing the recognition rate; if there is no match, the left camera of the binocular camera is continued to be used to collect the first image. If there is a match, the first feature point in the first image feature is extracted; finally, the first feature point is stereo matched with the second feature point to obtain the matching feature point; after the binocular camera is calibrated, three-dimensional reconstruction is performed based on the matching feature points to obtain the binocular visual image.

[0091] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by using the left camera and the right camera of the binocular camera to respectively collect images, using the SIFT and Hu feature fusion algorithm to determine whether the first image feature matches the reference feature, and obtaining the matching feature points according to the matching results, and finally performing three-dimensional reconstruction to obtain the binocular visual image, the accuracy of the binocular visual image acquisition can be guaranteed.

[0092] In one embodiment, performing stereo matching on the first feature point and the second feature point to obtain matching feature points includes:

[0093] Selecting a target feature point from the first feature points, and selecting a sub-window on the area where the target feature point is located;

[0094] According to the set similarity determination method, a target sub-window to be determined of several target feature points to be determined in the second feature point is determined. If the target sub-window to be determined has the largest similarity, the target sub-window to be determined is used as the target selection window, and the corresponding feature points summarized in the target selection window are used as matching feature points.

[0095] The working principle of the above technical solution is: in the process of matching the feature points of the flange, due to the limitations of object distance, field of view, etc., occlusion, tilt, etc. will always occur. In this regard, in the process of stereo matching, it is necessary to properly select feature points for stereo matching. The present invention selects target feature points from the first feature points, and selects a sub-window on the field where the target feature points are located; according to the set similarity judgment method, a number of target sub-windows to be determined of target feature points to be determined in the second feature points are determined. If the similarity of the target sub-window to be determined is the largest, the target sub-window to be determined is used as the target selection window, and the corresponding feature points summarized in the target selection window are used as matching feature points.

[0096] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by performing stereo matching, matching feature points can be accurately obtained in real time according to the situation where the edge features of the flange are more obvious.

[0097] In one embodiment, after the binocular camera is calibrated, three-dimensional reconstruction is performed based on matching feature points to obtain a binocular visual image, including:

[0098] The binocular camera is pre-calibrated using Zhang Zhengyou calibration method to obtain the camera's internal and external parameters;

[0099] According to the matching feature points, the local point cloud information is calculated, and the three-dimensional coordinate information of the flange to be measured is restored by point cloud stitching;

[0100] Based on the three-dimensional coordinate information, a binocular vision image is constructed and generated.

[0101] The working principle of the above technical solution is as follows: 3D reconstruction is also called parallax ranging method, which uses a binocular camera (left and right eyes) to capture images of the same scene and calculates the distance of the object by comparing the pixel information in the two images; the 3D coordinates of the object can be known according to the binocular camera imaging model and its formula, and the position of the binocular camera is taken as the origin, and the longitudinal distance of the workpiece from the camera is set, and the 3D coordinates of the image can be calculated. The present invention first uses Zhang Zhengyou calibration method to pre-calibrate the binocular camera to obtain the internal and external parameters of the camera; then, according to the matching feature points, the local point cloud information is calculated, and the 3D coordinate information of the flange to be measured is restored by point cloud splicing; finally, according to the 3D coordinate information, a binocular visual image is constructed and generated.

[0102] The beneficial effect of the above technical solution is: by adopting the solution provided by this embodiment, accurate binocular visual images can be obtained by calculating and obtaining three-dimensional coordinate information.

[0103] In one embodiment, based on the binocular visual image, a binocular camera measurement model is used to obtain a depth distance and a baseline distance of the binocular camera, including:

[0104] Extract feature points from the binocular vision image to obtain a feature point set;

[0105] A binocular camera measurement model is established based on the correspondence between feature point sets, and depth information and camera posture are calculated;

[0106] According to the depth information and camera pose, the depth distance and baseline distance are obtained.

[0107] The working principle of the above technical solution is: in order to obtain the depth distance and baseline distance of the binocular camera based on the binocular vision image and using the binocular camera measurement model, the present invention first extracts the feature points in the binocular vision image to obtain a feature point set; then establishes the binocular camera measurement model according to the correspondence between the feature point set, and calculates the depth information and camera posture; finally, obtains the depth distance and baseline distance according to the depth information and the camera posture.

[0108] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by extracting feature points in the binocular vision image and calculating the depth information and camera posture, accurate depth distance and baseline distance can be obtained.

[0109] In one embodiment, the binocular camera spacing adjustment is implemented based on the depth distance and the baseline distance, including:

[0110] Based on the depth distance and the baseline distance, a distance data set is established;

[0111] Based on the distance data set, a matching relationship library between the model and parameter data of the flange to be measured and the distance data is constructed;

[0112] According to the model and parameter data of the flange to be measured, the matching distance data is obtained by using the matching correspondence library;

[0113] Adjust the binocular camera distance based on the distance data.

[0114] The working principle of the above technical solution is: in order to realize the adjustment of the binocular camera spacing based on the depth distance and the baseline distance, the present invention first establishes a distance data set based on the depth distance and the baseline distance; then, based on the distance data set, constructs a matching correspondence library of the model and parameter data of the flange to be measured and the distance data; according to the model and parameter data of the flange to be measured, the matching correspondence library is used to obtain the matching distance data; finally, the binocular camera spacing is adjusted according to the distance data.

[0115] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by constructing a matching correspondence library between the model and parameter data of the flange to be measured and the distance data, and based on the matching correspondence library, obtaining matching distance data, the accuracy of adjusting the binocular camera spacing can be guaranteed.

[0116] In one embodiment, based on the depth distance and the baseline distance, the binocular camera spacing adjustment is implemented, and a binocular camera spacing adjustment control model is constructed to efficiently and accurately adjust the binocular camera spacing, specifically:

[0117] Based on the distance data, according to the measurement order of the flange to be measured, the adjustment difference of each distance adjustment is obtained;

[0118] According to the adjustment difference, obtain the left camera adjustment difference and the right camera adjustment difference;

[0119] Construct a control mechanism and a processor connected to the binocular camera; the control mechanism is used to control and adjust the distance of the left camera and the distance of the right camera, and the processor is used to generate control commands and control the control mechanism;

[0120] Using the left camera adjustment difference and the right camera adjustment difference, a corresponding control command set is constructed;

[0121] Using the processor, according to the control command set, a binocular camera distance adjustment control model is constructed to control the control mechanism and adjust the binocular camera distance to meet the measurement task requirements of the flange to be measured. Among them, the binocular camera distance adjustment control model is generated based on artificial neural network training.

[0122] The working principle of the above technical solution is: in order to achieve efficient and accurate adjustment of the binocular camera spacing, the present invention also constructs a binocular camera spacing adjustment control model, and the specific steps are: first, based on the distance data, according to the measurement order of the flange to be measured, the adjustment difference of each distance adjustment is obtained; then, according to the adjustment difference, the left camera adjustment difference and the right camera adjustment difference are obtained; a control mechanism and a processor connected to the binocular cameras are constructed; the control mechanism is used to control and adjust the left camera distance and the right camera distance, and the processor is used to generate control commands and control the control mechanism; finally, using the left camera adjustment difference and the right camera adjustment difference, a corresponding control command set is constructed; using the processor, according to the control command set, a binocular camera spacing adjustment control model is constructed, the control mechanism is controlled, and the binocular camera spacing is adjusted to meet the measurement task requirements of the flange to be measured; wherein, the binocular camera spacing adjustment control model is generated based on artificial neural network training.

[0123] The beneficial effect of the above technical solution is: by adopting the solution provided in this embodiment, by constructing a binocular camera distance adjustment control model, and according to the control mechanism and processor connected to the binocular camera, the binocular camera distance is adjusted according to the control command set, which can meet the measurement task requirements of the flange to be measured and improve the intelligence level of the control adjustment.

[0124] Binocular camera spacing adjustment system for flange measurement, such as Figure 3 As shown, including:

[0125] A measurement task data set acquisition module is used to acquire a measurement task data set of a flange to be measured;

[0126] A binocular vision image acquisition module is used to acquire a binocular vision image of the flange to be measured based on a binocular camera and a measurement task data set;

[0127] A depth distance and baseline distance acquisition module is used to obtain the depth distance and baseline distance of the binocular camera based on the binocular vision image and using the binocular camera measurement model;

[0128] The binocular camera distance adjustment implementation module is used to adjust the binocular camera distance based on the depth distance and the baseline distance.

[0129] The working principle of the above technical solution is: in order to realize the binocular camera spacing adjustment system for flange measurement, the present invention sets a measurement task data set acquisition module for acquiring the measurement task data set of the flange to be measured; sets a binocular vision image acquisition module for acquiring the binocular vision image of the flange to be measured based on the binocular camera according to the measurement task data set; sets a depth distance and baseline distance acquisition module for acquiring the depth distance and baseline distance of the binocular camera based on the binocular vision image and using the binocular camera measurement model; sets a binocular camera spacing adjustment implementation module for realizing binocular camera spacing adjustment based on the depth distance and baseline distance.

[0130] The beneficial effects of the above technical solution are: by adopting the solution provided in this embodiment, based on the measurement task data set of the flange to be measured, a binocular visual image of the flange to be measured is obtained based on the binocular camera, and by analyzing the binocular visual image, the depth distance and baseline distance of the binocular camera are obtained, and based on the depth distance and baseline distance, the binocular camera spacing is automatically adjusted, which can reduce manual adjustment errors and improve the efficiency and quality of binocular camera spacing adjustment.

[0131] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A binocular camera spacing adjustment method for flange measurement, characterized in that: include: Obtaining a measurement task data set of the flange to be measured; Based on the binocular camera and the measurement task data set, the binocular vision image of the flange to be measured is obtained; Based on the binocular vision image, the binocular camera measurement model is used to obtain the depth distance and baseline distance of the binocular camera; Based on the depth distance and baseline distance, the binocular camera distance adjustment is achieved.

2. The binocular camera spacing adjustment method for flange measurement according to claim 1, characterized in that: Get the measurement task data set of the flange to be measured, including: Obtain the model and parameter data of the flange to be measured; According to the model and parameter data, based on the set measurement task process, a measurement task data set is generated; the measurement task data set is used to measure the flange to be measured according to the set test sequence.

3. The binocular camera spacing adjustment method for flange measurement according to claim 2, characterized in that: According to the model and parameter data, based on the set measurement task process, a measurement task data set is generated, including: According to the model and parameter data, obtain the key words of the flange to be measured; Import keywords into the pre-set keyword analysis model and train to generate keyword vectors; Based on the keyword vectors, the K-means algorithm is used to perform cluster analysis to obtain several word vector groups, and multiple groups of measurement task items are generated based on the word vector groups; Generate a first measurement ranking according to the measurement priority of the measurement task item; Generate a second measurement sorting according to the binocular camera spacing adjustment amplitude value corresponding to the measurement task item; After weighted accumulation of the first measurement ranking and the second measurement ranking, a third measurement ranking is generated; Based on the third measurement sequence, a measurement task process is set, and a measurement task data set arranged according to the third measurement sequence is generated.

4. The binocular camera spacing adjustment method for flange measurement according to claim 1, characterized in that: Based on the binocular camera and the measurement task data set, the binocular vision image of the flange to be measured is obtained, including: Using the left camera of the binocular camera to capture a first image, preprocessing the first image to obtain a preprocessed first image, and extracting features from the preprocessed first image to obtain features of the first image; Using the right camera of the binocular camera to capture a second image, preprocessing the second image to obtain a preprocessed second image, performing feature extraction on the preprocessed second image to obtain second image features, and extracting second feature points from the second image features; Based on the set image feature library, the reference image feature is extracted, and based on the SIFT and Hu feature fusion algorithm, it is determined whether the first image feature matches the reference feature; if not, the left camera of the binocular camera is used to capture the first image; if they match, the first feature point in the first image feature is extracted; Perform stereo matching on the first feature point and the second feature point to obtain matching feature points; After the binocular camera is calibrated, three-dimensional reconstruction is performed based on the matching feature points to obtain the binocular visual image.

5. The binocular camera spacing adjustment method for flange measurement according to claim 4 is characterized in that: Performing stereo matching on the first feature point and the second feature point to obtain matching feature points includes: Selecting a target feature point from the first feature points, and selecting a sub-window on the area where the target feature point is located; According to the set similarity determination method, a target sub-window to be determined of several target feature points to be determined in the second feature point is determined. If the target sub-window to be determined has the largest similarity, the target sub-window to be determined is used as the target selection window, and the corresponding feature points summarized in the target selection window are used as matching feature points.

6. The binocular camera spacing adjustment method for flange measurement according to claim 4, characterized in that: After the binocular camera is calibrated, 3D reconstruction is performed based on matching feature points to obtain the binocular visual image, including: The binocular camera is pre-calibrated using Zhang Zhengyou calibration method to obtain the camera's internal and external parameters; According to the matching feature points, the local point cloud information is calculated, and the three-dimensional coordinate information of the flange to be measured is restored by point cloud stitching; Based on the three-dimensional coordinate information, a binocular vision image is constructed and generated.

7. The binocular camera spacing adjustment method for flange measurement according to claim 1, characterized in that: Based on the binocular vision image, the binocular camera measurement model is used to obtain the depth distance and baseline distance of the binocular camera, including: Extract feature points from the binocular vision image to obtain a feature point set; A binocular camera measurement model is established based on the correspondence between feature point sets, and depth information and camera posture are calculated; According to the depth information and camera pose, the depth distance and baseline distance are obtained.

8. The binocular camera spacing adjustment method for flange measurement according to claim 2, characterized in that: Based on the depth distance and baseline distance, the binocular camera spacing is adjusted, including: Based on the depth distance and the baseline distance, a distance data set is established; Based on the distance data set, a matching relationship library between the model and parameter data of the flange to be measured and the distance data is constructed; According to the model and parameter data of the flange to be measured, the matching distance data is obtained by using the matching correspondence library; Adjust the binocular camera distance based on the distance data.

9. The binocular camera spacing adjustment method for flange measurement according to claim 2, characterized in that: Based on the depth distance and the baseline distance, the binocular camera spacing adjustment is realized, and a binocular camera spacing adjustment control model is constructed to efficiently and accurately adjust the binocular camera spacing, specifically: Based on the distance data, according to the measurement order of the flange to be measured, the adjustment difference of each distance adjustment is obtained; According to the adjustment difference, obtain the left camera adjustment difference and the right camera adjustment difference; Construct a control mechanism and a processor connected to the binocular camera; the control mechanism is used to control and adjust the distance of the left camera and the distance of the right camera, and the processor is used to generate control commands and control the control mechanism; Using the left camera adjustment difference and the right camera adjustment difference, a corresponding control command set is constructed; Using the processor, according to the control command set, a binocular camera distance adjustment control model is constructed to control the control mechanism and adjust the binocular camera distance to meet the measurement task requirements of the flange to be measured. Among them, the binocular camera distance adjustment control model is generated based on artificial neural network training.

10. A binocular camera spacing adjustment system for flange measurement, characterized in that: include: A measurement task data set acquisition module is used to acquire a measurement task data set of a flange to be measured; A binocular vision image acquisition module is used to acquire a binocular vision image of the flange to be measured based on a binocular camera and a measurement task data set; A depth distance and baseline distance acquisition module is used to obtain the depth distance and baseline distance of the binocular camera based on the binocular vision image and using the binocular camera measurement model; The binocular camera distance adjustment implementation module is used to adjust the binocular camera distance based on the depth distance and the baseline distance.

Images