Depth image-based position and posture adjustment method for reinforcing steel bar binding robot

Abstract

The invention provides a depth image-based position and posture adjustment method for a reinforcing steel bar binding robot. In the method, firstly, a depth image of a reinforcement cage is obtained; then feature extraction is carried out on the depth image to obtain feature points; then, pose information of the reinforcement cage is obtained through calculation based on the feature points, and coordinate information of the target position is calculated based on the pose information; and based on the current coordinates and the coordinate information of the reinforcing steel bar binding robot, planning an operation path of the reinforcing steel bar binding robot, and based on the operation path, controlling the reinforcing steel bar binding robot to move to a target position. And when the steel bar binding robot reaches the target position, the pose of the steel bar binding robot is adjusted based on the pose information. Therefore, by means of the method, the steel bar binding robot can autonomously move, adjust the posture and quickly reach the working position to work, manual operation is not needed, and the working efficiency and quality are improved. The problem that in the prior art, the operation efficiency of a steel bar binding robot is low is solved.

Description

Technical Field

[0001] The present application relates to the field of automation technology, and in particular to a method for adjusting the posture of a steel bar binding robot based on a depth image. Background Art

[0002] In the manufacturing of precast concrete components for bridges, the tying of steel cages is an important and complex process. In order to improve the efficiency and quality of tying, the industry has developed a four-wheeled steel cage tying robot system for steel cage tying. The initial posture of the steel cage tying robot before reaching the steel cage must be parallel to the spatial posture of the surface to be tied, so as to ensure that it always remains parallel to the steel cage during the subsequent tying process and ensure the accuracy of the tying position.

[0003] In the prior art, the position and posture of the steel bar tying robot are usually adjusted manually by remote control equipment so that it reaches the position of the steel bar cage and remains parallel to the steel bar cage for operation. However, this method requires a lot of human resources, and manual operation is prone to errors, resulting in low efficiency of the steel bar tying robot. Summary of the invention

[0004] In view of this, the present application provides a posture adjustment method of a steel bar tying robot based on a depth image to solve the problem of low operating efficiency of the steel bar tying robot in the prior art.

[0005] To achieve the above objectives, this application provides the following technical solutions:

[0006] The first aspect of the present application discloses a method for adjusting the posture of a steel bar tying robot based on a depth image, comprising:

[0007] Acquire a depth image of the steel cage;

[0008] Extracting features from the depth image to obtain feature points; wherein the feature points are used to characterize local features of the steel cage;

[0009] The posture information of the steel cage is calculated based on the feature points; wherein the posture information includes a plane normal vector, position information and direction information;

[0010] Calculating the coordinate information of the target position based on the posture information; wherein the target position is the working position that the steel bar tying robot needs to reach;

[0011] Based on the current coordinates of the steel bar tying robot and the coordinate information, planning the operation path of the steel bar tying robot, and controlling the steel bar tying robot to move to the target position based on the operation path;

[0012] When the steel bar tying robot reaches the target position, the posture of the steel bar tying robot is adjusted based on the posture information.

[0013] Optionally, in the above method, extracting features from the depth image to obtain feature points includes:

[0014] Performing Gaussian blur processing on the depth image machine by using SIFT algorithm to generate an image pyramid of the depth image;

[0015] For each scale space of the image pyramid, the extreme point of the current scale space is used as the feature point, and a feature descriptor of the feature point is generated.

[0016] Optionally, the above method, after extracting features from the depth image to obtain feature points, further includes:

[0017] The feature points are matched with the feature points in the pre-built feature point data set, and the successfully matched feature points are screened out.

[0018] Optionally, in the above method, the step of performing feature matching between the feature points and feature points in a pre-constructed feature point data set includes:

[0019] Build the best model for feature matching through the RANSAC algorithm;

[0020] The optimal model is used to perform feature matching between the feature points and feature points in a pre-constructed feature point data set, and successfully matched feature points are screened out.

[0021] Optionally, in the above method, constructing the best model for feature matching by using the RANSAC algorithm includes:

[0022] Randomly select a subset from the feature point data set;

[0023] evaluating an initial model of feature matching based on the subset;

[0024] Evaluate the degree of fit between the feature points in the feature point data set except the subset and the model, and generate an inlier set; wherein the feature points with a high degree of fit are inliers, and the feature points with a low degree of fit are outliers;

[0025] If the number of feature points in the interior point set is greater than a preset threshold, determining the initial model as the optimal model;

[0026] If the number of feature points in the inner point set is not greater than the threshold, the process returns to executing the random selection of a subset from the feature point data set until the number of feature points in the inner point set is greater than the preset threshold, and the initial model is determined to be the optimal model.

[0027] Optionally, in the above method, the calculating the position information of the steel cage based on the feature points includes:

[0028] Converting the feature points into three-dimensional point cloud data;

[0029] Performing plane fitting on the three-dimensional point cloud data to calculate and obtain the initial position information of the steel cage;

[0030] The initial position and posture information is globally optimized using a beam adjustment algorithm to obtain the position and posture information.

[0031] Optionally, in the above method, performing plane fitting on the three-dimensional point cloud data to calculate the initial position information of the steel cage includes:

[0032] Identifying a long plane on the side of the steel cage;

[0033] Using a principal component analysis algorithm, plane fitting is performed on the three-dimensional point cloud data to obtain the normal vector and position information of the long plane;

[0034] Based on the normal vector and the position information, the initial position information of the steel cage is calculated.

[0035] Optionally, in the above method, controlling the steel bar tying robot to move to the target position based on the operation path includes:

[0036] When the steel bar tying robot moves, a depth image of the environment on the operation path is collected;

[0037] Determine whether there is an obstacle on the working path based on the environmental depth image;

[0038] If there is no obstacle, controlling the steel bar tying robot to move to the target position based on the operation path;

[0039] If there is an obstacle, the operation path is adjusted to generate a new operation path, and the steel bar binding robot is controlled to move to the target position according to the new operation path.

[0040] Optionally, in the above method, if an obstacle exists, adjusting the operation path to generate a new operation path includes:

[0041] If there is an obstacle, the coordinates of the steel bar tying robot and the obstacle are obtained;

[0042] The obstacle coordinates are set as non-way points, and based on the coordinate information of the steel bar tying robot, the obstacle coordinates and the target position, an operation path that does not include the non-way points is regenerated.

[0043] It can be seen from the above technical scheme that in a method for adjusting the posture of a steel bar binding robot based on a depth image provided by the present application, a depth image of a steel bar cage is first obtained. Then, feature extraction is performed on the depth image to obtain feature points. Then, the posture information of the steel bar cage is calculated based on the feature points, and the coordinate information of the target position is calculated based on the posture information. Based on the current coordinates and coordinate information of the steel bar binding robot, the operation path of the steel bar binding robot is planned, and the steel bar binding robot is controlled to move to the target position based on the operation path. When the steel bar binding robot reaches the target position, the posture of the steel bar binding robot is adjusted based on the posture information. It can be seen that the present application obtains a depth image of the steel bar cage and extracts feature points, and performs posture recognition on the steel bar cage based on the feature points, so as to ensure that the steel bar binding robot remains parallel to the surface to be bound of the steel bar cage during operation, thereby improving the accuracy of the binding position. At the same time, the steel bar binding robot can move and adjust its posture autonomously, quickly reach the working position and perform operations without manual operation, thereby improving work efficiency and quality. The problem of low operating efficiency of the steel bar binding robot in the prior art is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0045] Figure 1 A flowchart of a method for adjusting the posture of a steel bar tying robot based on a depth image disclosed in an embodiment of the present application;

[0046] Figure 2 A schematic diagram of an operation scene of a steel bar tying robot disclosed in an embodiment of the present application;

[0047] Figure 3 This is an example diagram of generating a feature descriptor disclosed in an embodiment of the present application;

[0048] Figure 4 This is a flowchart of an implementation of step S103 disclosed in an embodiment of the present application;

[0049] Figure 5 This is a flowchart of an implementation method of controlling the rebar tying robot to move to the target position based on the operation path in step S105 disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0052] Furthermore, in this document, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0053] As can be seen from the background technology, in the prior art, the position and posture of the steel bar tying robot are usually manually adjusted by remote control through a remote control device so that it reaches the position of the steel bar cage and remains parallel to the steel bar cage for operation. However, this method requires a lot of human resources, and manual operation is prone to errors, resulting in low operating efficiency of the steel bar tying robot.

[0054] In view of this, the present application provides a method for adjusting the posture of a steel bar tying robot based on a depth image to solve the problem of low operating efficiency of the steel bar tying robot in the prior art.

[0055] The embodiment of the present application provides a method for adjusting the posture of a steel bar tying robot based on a depth image. Figure 1 As shown, specifically including:

[0056] S101, obtaining a depth image of the steel cage.

[0057] It should be noted that the depth camera is first used to conduct a comprehensive scan of the current working environment. Figure 2As shown. The two lenses of the depth camera provide parallax information, and the depth image is generated by calculating the parallax value of each pixel, so as to obtain the depth image of the steel cage in the current working environment. Among them, the depth camera measures the distance of the object by emitting infrared light and capturing the time difference (Time of Flight, ToF principle) when it is reflected back, thereby obtaining three-dimensional depth information, so that the robot can better understand its spatial environment. In this example, the depth camera uses a model with a resolution of 640x480 pixels and a frame rate of 30 frames per second, a depth range of 0.5-5 meters, and a field of view of 70 degrees horizontally and 50 degrees vertically to ensure sufficient scanning coverage and accuracy.

[0058] It should also be noted that the method of the present application can be applied to the Robot Operating System (ROS), which is a flexible robot software development framework that provides a set of tools and libraries to help developers build robot applications. It supports the integration of multiple sensors and actuators, as well as advanced functions such as simulation, visual recognition, and path planning. For the development of AGVs and rebar tying robots, ROS provides a powerful platform to implement complex navigation and task execution strategies.

[0059] S102, extracting features from the depth image to obtain feature points; wherein the feature points are used to characterize local features of the steel cage.

[0060] It should be noted that after obtaining the depth image of the steel cage, the SIFT (Scale Invariant Feature Transform) algorithm is used to extract the features of the depth image for each steel cage to obtain the feature points of the depth image, which are used to characterize the local features of the steel cage. The SIFT algorithm first generates image pyramids of different scales through Gaussian blur, and then searches for extreme points in each scale space as feature points. The gradient direction is calculated around each feature point to generate a feature descriptor, such as Figure 3 The advantage of SIFT is that it is invariant to scale and rotation, and can extract stable feature points in complex environments. The feature extraction algorithm can extract stable and recognizable key points from the image. These key points are still robust under different viewing angles and lighting conditions.

[0061] In other embodiments of the present application, feature extraction may be performed on the depth image using an ORB (Oriented Rapid and Rotation Invariant Binary Feature Description) algorithm to obtain feature points of the depth image.

[0062] Optionally, in another embodiment of the present application, after executing step S102, the following may also be included:

[0063] The feature points are matched with the feature points in the pre-built feature point data set, and the successfully matched feature points are screened out.

[0064] It should be noted that after extracting the feature points in the depth image, in order to effectively eliminate the erroneous feature points and improve the accuracy of subsequent calculations, the feature points are matched with the feature points in the pre-built feature point data set to screen out the successfully matched feature points. Among them, the feature point data set contains the real feature points that characterize the local features of the steel cage.

[0065] Optionally, in another embodiment of the present application, the above step of performing feature matching between the feature points and the feature points in the pre-built feature point data set and screening out the successfully matched feature points may include:

[0066] Build the best model for feature matching through the RANSAC algorithm;

[0067] The optimal model is used to perform feature matching between the feature points and feature points in a pre-constructed feature point data set, and successfully matched feature points are screened out.

[0068] It should be noted that, in this embodiment, the RANSAC (Random Sampling Consistency) algorithm is used for feature matching to ensure the accuracy and stability of the matching. The RANSAC algorithm is an algorithm for parameter estimation. It finds the internal point set through random sampling and removes outliers for model fitting. The RANSAC algorithm finds the best model for feature matching in a data set containing noise and external points through random sampling and iterative optimization. Then, using the best model, the feature points and the feature points in the pre-built feature point data set are input into the best model for feature matching, and the successfully matched feature points are screened out. If the current feature point successfully matches a feature point in the feature point data set, it means that the current feature point is an accurate feature point. If the current feature point fails to match every feature point in the feature point data set, it means that the current feature point may be an incorrect feature point. Then the successfully matched feature points are screened out for retention, and the unsuccessful matching feature points are removed to provide data accuracy.

[0069] Optionally, in another embodiment of the present application, the optimal model for feature matching constructed by the RANSAC algorithm may include:

[0070] A subset is randomly selected from the feature point dataset.

[0071] An initial model of feature matching is evaluated based on the subset.

[0072] The degree of fit between the feature points in the feature point data set except the subset and the model is evaluated, and an inlier set is generated; wherein the feature points with a high degree of fit are inliers, and the feature points with a low degree of fit are outliers.

[0073] If the number of feature points in the interior point set is greater than a preset threshold, the initial model is determined to be the optimal model.

[0074] If the number of feature points in the inner point set is not greater than the threshold, the process returns to executing the random selection of a subset from the feature point data set until the number of feature points in the inner point set is greater than the preset threshold, and the initial model is determined to be the optimal model.

[0075] It should be noted that, first, a subset is randomly selected from the feature point data set to estimate the initial matching model; then, the initial model of feature matching is evaluated based on the subset. The degree of fit between the feature points in the feature point data set except the subset and the model is evaluated, and the inliers that meet the model are added to the inliers set, wherein the feature points with a high degree of fit are inliers, and the feature points with a low degree of fit are outliers; if the size of the inliers set exceeds a predetermined threshold, the model is considered to be the best model; otherwise, the above steps are repeated until the best model is found or the upper limit of the number of iterations is reached.

[0076] S103, calculating the position and posture information of the steel cage based on the feature points; wherein the position and posture information includes a plane normal vector, position information and direction information.

[0077] It should be noted that after the feature points that can characterize the specific features of the steel cage are extracted, the posture information of the steel cage is calculated based on the feature points; wherein the posture information includes a plane normal vector, position information and direction information.

[0078] Optionally, in another embodiment of the present application, an implementation of step S103 is as follows: Figure 4 As shown, it may include:

[0079] S401, converting feature points into three-dimensional point cloud data.

[0080] It should be noted that the side of the steel cage has obvious geometric features, such as long planes and evenly arranged steel bars, which can be used for identification and positioning. The feature points are converted into 3D point cloud data, and the geometric features (long planes) of the side of the steel cage are identified in the 3D point cloud data. Among them, the 3D point cloud data is a collection of a large number of points, each of which has its own 3D coordinates (X, Y, Z), which are used to represent the 3D shape of an object or environment.

[0081] S402, performing plane fitting on the three-dimensional point cloud data to calculate and obtain initial position information of the steel cage.

[0082] It should be noted that the long plane on the side of the steel cage is first identified, and then the three-dimensional point cloud data is plane-fitted by the principal component analysis (PCA) algorithm to calculate the normal vector and position information of the long plane on the side of the steel cage, wherein the principal component analysis (PCA) algorithm is a statistical method for dimensionality reduction and data compression, and the principal component is determined by finding the direction of the maximum variance in the data. Finally, the normal vector and the position information are used to calculate the initial posture information of the steel cage, including the position and direction of the side plane of the steel cage.

[0083] S403, using a beam adjustment algorithm to globally optimize the initial posture information to obtain posture information.

[0084] It should be noted that the initial posture is globally optimized using the Bundle Adjustment algorithm to obtain the optimized posture, including the normal vector, position and direction of the side plane of the steel cage.

[0085] S104. Calculate the coordinate information of the target position based on the posture information; wherein the target position is the working position that the steel bar tying robot needs to reach.

[0086] It should be noted that the position coordinates of the steel cage can be obtained based on the position information in the posture information ( , ), based on the normal vector in the pose information, the distance D between the steel bar tying robot and the steel cage can be obtained, and based on the direction information in the pose information, the direction angle of the side of the steel cage can be obtained. Then, the coordinate information of the target position is calculated using the following formula:

[0087]

[0088]

[0089] in,( , ) are the coordinates of the target position, and the target position is the working position that the steel bar tying robot needs to reach.

[0090] S105. Based on the current coordinates and coordinate information of the steel bar tying robot, plan the operation path of the steel bar tying robot, and control the steel bar tying robot to move to the target position based on the operation path.

[0091] It should be noted that the coordinates of the current position of the steel bar tying robot are obtained, and then based on the current coordinates and coordinate information of the steel bar tying robot, the A* path planning algorithm is used to plan the shortest working path from the steel bar tying robot to the steel cage. In this process, the algorithm will consider the geometric constraints of the environment (such as walls, columns, etc.) and dynamic constraints (such as moving workers). Then the steel bar tying robot is controlled to move to the target position based on the working path.

[0092] Optionally, in another embodiment of the present application, in step S105, a steel bar tying robot is controlled to move to a target position based on the operation path, such as Figure 5 As shown, it may include:

[0093] S501. When the steel bar tying robot moves, it collects a depth image of the environment on the working path.

[0094] It should be noted that during the movement of the steel bar tying robot, the depth camera continues to be used to scan the surrounding environment so as to detect new obstacles or environmental changes in a timely manner.

[0095] S502: Determine whether there is an obstacle on the working path based on the environmental depth image.

[0096] S503: If there is no obstacle, the steel bar tying robot is controlled to move to the target position based on the operation path.

[0097] S504: If there is an obstacle, adjust the operation path, generate a new operation path, and control the steel bar binding robot to move to the target position according to the new operation path.

[0098] It should be noted that if there is an obstacle, the operation path is adjusted, a new operation path is generated, and the steel bar tying robot is controlled to move to the target position according to the new operation path. In this embodiment, the dynamic window approach (DWA) is used to dynamically adjust the path according to the results of real-time environmental perception to avoid obstacles that are about to collide. During the entire movement process, the control system needs to continuously adjust the speed and steering of the steel bar tying robot to adapt to changes in path planning and obstacle avoidance needs.

[0099] Optionally, in another embodiment of the present application, if there is an obstacle in the above step S504, an implementation method of adjusting the operation path and generating a new operation path may include:

[0100] If there is an obstacle, the coordinates of the steel bar tying robot and the obstacle are obtained;

[0101] The obstacle coordinates are set as non-way points, and based on the coordinate information of the steel bar tying robot, the obstacle coordinates and the target position, an operation path that does not include the non-way points is regenerated.

[0102] It should be noted that if there is an obstacle, the coordinates of the rebar tying robot and the obstacle coordinates are obtained, and the obstacle coordinates are set as a non-way point. The coordinates of the rebar tying robot are used as the starting point, the coordinate information of the target position is used as the end point, and the obstacle coordinates are used as the non-way point, and the operation path that does not include the non-way point is regenerated.

[0103] S106. When the steel bar tying robot reaches the target position, the posture of the steel bar tying robot is adjusted based on the posture information.

[0104] It should be noted that when the rebar tying robot reaches the target position, the posture of the rebar tying robot is adjusted based on the posture information of the rebar cage to ensure that the spatial posture of the rebar tying robot is parallel to the surface of the rebar cage to be tied (such as the side of the rebar cage mentioned above), and then the rebar tying robot can perform the rebar cage tying operation.

[0105] In a method for adjusting the posture of a steel bar binding robot based on a depth image provided by an embodiment of the present application, a depth image of a steel bar cage is first obtained. Then, feature extraction is performed on the depth image to obtain feature points. Then, the posture information of the steel bar cage is calculated based on the feature points, and the coordinate information of the target position is calculated based on the posture information. Based on the current coordinates and coordinate information of the steel bar binding robot, the operation path of the steel bar binding robot is planned, and the steel bar binding robot is controlled to move to the target position based on the operation path. When the steel bar binding robot reaches the target position, the posture of the steel bar binding robot is adjusted based on the posture information. It can be seen that the present application obtains a depth image of the steel bar cage and extracts feature points, and performs posture recognition on the steel bar cage based on the feature points, so as to ensure that the steel bar binding robot remains parallel to the surface to be bound of the steel bar cage during operation, thereby improving the accuracy of the binding position. At the same time, the steel bar binding robot can move and adjust its posture autonomously, quickly reach the working position and perform operations without manual operation, thereby improving work efficiency and quality. The problem of low operating efficiency of the steel bar binding robot in the prior art is solved.

[0106] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0107] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0108] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for adjusting the posture of a steel bar tying robot based on a depth image, characterized in that: include: Acquire a depth image of the steel cage; Performing feature extraction on the depth image to obtain feature points; wherein the feature points are used to characterize local features of the steel cage; The posture information of the steel cage is calculated based on the feature points; wherein the posture information includes a plane normal vector, position information and direction information; Calculating the coordinate information of the target position based on the posture information; wherein the target position is the working position that the steel bar tying robot needs to reach; Based on the current coordinates of the steel bar tying robot and the coordinate information, planning the operation path of the steel bar tying robot, and controlling the steel bar tying robot to move to the target position based on the operation path; When the steel bar tying robot reaches the target position, the posture of the steel bar tying robot is adjusted based on the posture information.

2. The method according to claim 1, characterized in that The extracting features from the depth image to obtain feature points includes: Performing Gaussian blur processing on the depth image machine by using SIFT algorithm to generate an image pyramid of the depth image; For each scale space of the image pyramid, the extreme point of the current scale space is used as the feature point, and a feature descriptor of the feature point is generated.

3. The method according to claim 1, characterized in that: After extracting features from the depth image to obtain feature points, the method further includes: The feature points are matched with the feature points in the pre-built feature point data set, and the successfully matched feature points are screened out.

4. The method according to claim 3, characterized in that The feature matching of the feature points with the feature points in the pre-built feature point data set includes: Build the best model for feature matching through the RANSAC algorithm; The optimal model is used to perform feature matching between the feature points and feature points in a pre-constructed feature point data set, and successfully matched feature points are screened out.

5. The method according to claim 4, characterized in that The optimal model for feature matching is constructed by the RANSAC algorithm, including: Randomly select a subset from the feature point data set; evaluating an initial model of feature matching based on the subset; Evaluate the degree of fit between the feature points in the feature point data set except the subset and the model, and generate an inlier set; wherein the feature points with a high degree of fit are inliers, and the feature points with a low degree of fit are outliers; If the number of feature points in the interior point set is greater than a preset threshold, determining the initial model as the optimal model; If the number of feature points in the inner point set is not greater than the threshold, the process returns to executing the random selection of a subset from the feature point data set until the number of feature points in the inner point set is greater than the preset threshold, and the initial model is determined to be the optimal model.

6. The method according to claim 1, characterized in that The step of calculating the position and posture information of the steel cage based on the feature points includes: Converting the feature points into three-dimensional point cloud data; Performing plane fitting on the three-dimensional point cloud data to calculate and obtain the initial position information of the steel cage; The initial position and posture information is globally optimized using a beam adjustment algorithm to obtain the position and posture information.

7. The method according to claim 6, characterized in that The performing plane fitting on the three-dimensional point cloud data to calculate the initial position information of the steel cage includes: Identifying a long plane on the side of the steel cage; Using a principal component analysis algorithm, plane fitting is performed on the three-dimensional point cloud data to obtain the normal vector and position information of the long plane; Based on the normal vector and the position information, the initial position information of the steel cage is calculated.

8. The method according to claim 1, characterized in that The step of controlling the steel bar tying robot to move to the target position based on the operation path includes: When the steel bar tying robot moves, a depth image of the environment on the operation path is collected; Determine whether there is an obstacle on the working path based on the environmental depth image; If there is no obstacle, controlling the steel bar tying robot to move to the target position based on the operation path; If there is an obstacle, the operation path is adjusted to generate a new operation path, and the steel bar binding robot is controlled to move to the target position according to the new operation path.

9. The method according to claim 8, characterized in that If there is an obstacle, adjusting the operation path to generate a new operation path includes: If there is an obstacle, the coordinates of the steel bar tying robot and the obstacle are obtained; The obstacle coordinates are set as non-way points, and based on the coordinate information of the steel bar tying robot, the obstacle coordinates and the target position, an operation path that does not include the non-way points is regenerated.

Images