Base coordinate system determination method, electronic equipment and medium

Abstract

The invention provides a base coordinate system determination method, electronic equipment and a medium, after a target three-dimensional model is imported into simulation software, target three-dimensional models corresponding to a plurality of specific poses are obtained based on a plurality of preset specific poses and a base coordinate system of the target three-dimensional model, and the target three-dimensional models are determined according to a three-dimensional plane in a simulation space. The target three-dimensional models corresponding to the multiple specific poses are simulated, multiple simulation results are obtained, finally, the simulation result corresponding to the target static posture is determined from the simulation results through a pre-trained detection model, and then the base coordinate system matched with the target static posture serves as the base coordinate system of the target simulation result. Therefore, the base coordinate system matched with the target simulation result can be determined under the condition that the base coordinate system does not need to be calibrated manually, and the working efficiency is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of three-dimensional modeling, and in particular to a method for determining a base coordinate system, an electronic device, and a medium. Background Art

[0002] At present, with the continuous advancement of artificial intelligence technology and the need for the integration of three-dimensional reconstruction services, simulation services based on three-dimensional models are becoming more and more widely used in people's lives.

[0003] A common simulation process is to import a 3D model of a target object into simulation software, and the simulation software performs 3D reconstruction based on the imported 3D model, determines the coordinates of the target object, and adjusts the coordinates of the target object to perform simulation.

[0004] However, when the simulation software performs 3D reconstruction on the 3D model of the target object, there is a situation where the reconstructed 3D model does not match the base coordinate system in which it is located. Based on this, the present application provides a method for determining the base coordinate system. Summary of the invention

[0005] In view of this, embodiments of the present application provide a base coordinate system determination method, electronic device, and medium, which can improve work efficiency.

[0006] In a first aspect, an embodiment of the present application provides a method for determining a base coordinate system, which includes: determining a target three-dimensional model; adjusting the position and posture of the target three-dimensional model according to multiple specific postures to obtain target three-dimensional models corresponding to the multiple specific postures; simulating the target three-dimensional models corresponding to the multiple specific postures according to the three-dimensional plane in the simulation space to obtain simulation results corresponding to the multiple specific postures, and for each specific posture, the simulation result of the specific posture includes a target three-dimensional model of the target three-dimensional model in the specific posture that is in a static posture after falling on the three-dimensional plane; determining, from each simulation result, a simulation result that matches the target static posture as the target simulation result by detecting the model, and determining the base coordinate system of the target simulation result according to the base coordinate system that matches the target static posture.

[0007] In the second aspect, an embodiment of the present application provides a base coordinate system determination device, which includes: an acquisition module for determining a target three-dimensional model; an adjustment module for adjusting the position and posture of the target three-dimensional model according to multiple specific postures, and obtaining target three-dimensional models corresponding to multiple specific postures; a simulation module for simulating the target three-dimensional models corresponding to multiple specific postures according to a three-dimensional plane in a simulation space, and obtaining simulation results corresponding to multiple specific postures, wherein for each specific posture, the simulation result of the specific posture includes a target three-dimensional model of the target three-dimensional model in the specific posture that is in a static posture after falling on the three-dimensional plane; a determination module for determining, from each simulation result, a simulation result that matches the target static posture as the target simulation result through a detection model, and determining the base coordinate system of the target simulation result according to the base coordinate system that matches the target static posture.

[0008] In a third aspect, an embodiment of the present application provides a robot, comprising a control module, wherein the control module is used to execute the base coordinate system determination method of the first aspect.

[0009] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising: a processor; and a memory for storing processor executable instructions, wherein the processor is used to execute the base coordinate system determination method of the first aspect above.

[0010] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the base coordinate system determination method of the first aspect described above.

[0011] In a sixth aspect, an embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor of a computer device, the computer device is enabled to execute the base coordinate system determination method of the first aspect described above.

[0012] In a seventh aspect, an embodiment of the present application provides a chip, comprising: a processor; and a memory for storing processor executable instructions, wherein the processor is used to execute the base coordinate system determination method of the first aspect described above.

[0013] The embodiments of the present application provide a method, an electronic device and a medium for determining a base coordinate system. After a target three-dimensional model is imported into simulation software, target three-dimensional models corresponding to multiple specific postures are obtained based on multiple preset specific postures and the base coordinate system of the target three-dimensional model. Then, according to the three-dimensional plane in the simulation space, the target three-dimensional models corresponding to the multiple specific postures are simulated to obtain multiple simulation results of the target three-dimensional model that may include the target three-dimensional model in a static posture after falling on the three-dimensional plane. Finally, by detecting the model, a target simulation result matching the target static posture is determined from each simulation result, and then the base coordinate system of the target simulation result is determined according to the base coordinate system matching the target static posture.

[0014] In this way, the embodiment of the present application can determine the target simulation result and the base coordinate system of the target simulation result that matches the target static posture from the target three-dimensional model corresponding to each specific posture and the simulation results of the three-dimensional plane in the simulation space through the detection model. Without the need for manual calibration of the base coordinate system, the base coordinate system that matches the target simulation result can be determined, thereby improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic diagram of a scene of a base coordinate system provided in an embodiment of the present application.

[0016] Figure 2 A schematic diagram of a scene of a base coordinate system provided in an embodiment of the present application.

[0017] Figure 3 Shown is a schematic diagram of the system architecture of a simulation system provided by an exemplary embodiment of the present application.

[0018] Figure 4 The figure is a schematic flow chart of a method for determining a base coordinate system provided by an exemplary embodiment of the present application.

[0019] Figure 5 This is a schematic diagram of the collection points corresponding to the same azimuth provided in this manual.

[0020] Figure 6 The figure is a schematic flow chart of a method for determining a base coordinate system provided by an exemplary embodiment of the present application.

[0021] Figure 7 Shown is a schematic structural diagram of a base coordinate system determination device provided by an exemplary embodiment of the present application.

[0022] Figure 8 FIG. 1 is a block diagram of an electronic device for executing a method for determining a base coordinate system according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0023] 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.

[0024] Application Overview

[0025] At present, when performing simulation, for each 3D model that needs to be imported into the simulation software, the 3D model usually has its corresponding base coordinate system. The base coordinate system is predefined when the 3D model is constructed. The base coordinate system can be used to represent the reference starting point of the 3D model in its own space, as well as the relative position relationship between the components in the 3D model, such as Figure 1 shown.

[0026] Figure 1 A scene schematic diagram of a base coordinate system provided in an embodiment of the present application. In the figure, the three-dimensional model corresponding to the target object is a cuboid as an example for explanation, and the base coordinate system of the cuboid can be the Cartesian coordinate system shown in the figure. The base coordinate system is predefined when constructing the cuboid. Based on the base coordinate system and the three-dimensional model, the simulation software can perform simulation.

[0027] Taking the simulation of placing the cuboid on a horizontal plane as an example, the three-dimensional model file corresponding to the cuboid can be imported into the simulation software, and the simulation software can perform three-dimensional reconstruction of the three-dimensional model of the cuboid. The three-dimensional model is reconstructed in the simulation space inside the simulation software, and based on the base coordinate system of the three-dimensional model, the component of the three-dimensional model in the Z-axis direction is adjusted to 0. In this way, the simulation process of placing the cuboid on a horizontal plane can be realized.

[0028] However, due to sensor errors, algorithm deviations, and other problems, the base coordinate system of the 3D model reconstructed in the simulation software is often different from the original base coordinate system. Figure 1 There are some deviations in the coordinate system shown in Figure 2 shown.

[0029] Figure 2 A schematic diagram of a scene of a base coordinate system provided in an embodiment of the present application. Figure 2 In the simulation space, the reconstructed 3D model does not match its base coordinate system. If the base coordinate system is matched with the 3D coordinate system in the simulation space, the posture of the reconstructed 3D model in the simulation space is the same as Figure 1 There must be a gap between the three-dimensional models in Figure 2If the posture shown is reconstructed in three-dimensional space, it will take a long time to calculate in both the horizontal and vertical directions to accurately adjust the position of the three-dimensional model in the three-dimensional space.

[0030] At the same time, when constructing a 3D model based on a base coordinate system, usually only the relative position relationship between the components in the 3D model is considered, so as to construct a relatively accurate model, but the matching relationship between the base coordinate system and the 3D model is often ignored. Therefore, the relative position relationship between the constructed 3D model and its base coordinate system often presents the following Figure 2 Based on this, a method for obtaining a base coordinate system that matches the three-dimensional model is urgently needed.

[0031] In response to the above technical problems, an embodiment of the present application provides a method for determining a base coordinate system. After the target three-dimensional model is imported into the simulation software, based on a plurality of preset specific postures and the base coordinate system of the target three-dimensional model, target three-dimensional models corresponding to the plurality of specific postures are obtained. Then, according to the three-dimensional plane in the simulation space, the target three-dimensional models corresponding to the plurality of specific postures are simulated to obtain a plurality of simulation results of the target three-dimensional model that may include the target three-dimensional model being in a static posture after falling on the three-dimensional plane. Finally, by detecting the model, a target simulation result matching the static posture of the target is determined from each simulation result, and then the base coordinate system of the target simulation result is determined according to the base coordinate system matching the static posture of the target.

[0032] In this way, the embodiment of the present application can determine the target simulation result and the base coordinate system of the target simulation result that matches the target static posture from the target three-dimensional model corresponding to each specific posture and the simulation results of the three-dimensional plane in the simulation space through the detection model. Without the need to manually calibrate the base coordinate system, the base coordinate system that matches the target simulation result can be determined.

[0033] Exemplary Systems

[0034] Figure 3 FIG. 1 is a schematic diagram of a system architecture of a simulation system 100 provided by an exemplary embodiment of the present application. Figure 3 As shown, the simulation system may include a terminal 110, which is used to import a target three-dimensional model 120 of a target object into a simulation space, and adjust the position and posture of the target three-dimensional model 120 and its base coordinate system to obtain the target three-dimensional model 120 and its base coordinate system that match the three-dimensional coordinate system of the simulation space. The terminal 110 may be a terminal held by a user, such as a mobile phone, a tablet computer, or a smart device, or may be a device such as a display screen panel provided by a service provider to a user.

[0035] The terminal 110 may be provided with a controller 111, and the controller 111 is used to adjust the posture of the target three-dimensional model 120 and simulate the target three-dimensional model 111 after the posture adjustment.

[0036] In one example, after the three-dimensional model of the target object is imported into the terminal 110, the controller 111 in the terminal 110 can determine the three-dimensional model of the target object as the target three-dimensional model 120. Therefore, the controller 111 can adjust the position and posture of the target three-dimensional model 120 according to a plurality of preset specific postures to obtain target three-dimensional models corresponding to a plurality of specific postures. Afterwards, the controller 111 can determine a three-dimensional plane in the simulation space, and based on the three-dimensional plane, perform a collision simulation test on the target three-dimensional models corresponding to the above-mentioned plurality of specific postures to obtain simulation results corresponding to each specific posture. Finally, by detecting the model, the target simulation result matching the target static posture is determined from each simulation result, and the base coordinate system of the target simulation result is determined based on the base coordinate system matching the target static posture.

[0037] Exemplarily, the controller 111 may include at least one of a logic controller, a machine vision controller, a motion controller, and the like.

[0038] In other examples, the controller 111 may be located outside the terminal 110 , such as being a device independent of the terminal 110 .

[0039] It should be understood that the above application scenario examples are only shown to facilitate understanding of the spirit and principle of the present application, and the embodiments of the present application are not limited thereto. On the contrary, the embodiments of the present application can be applied to any scenario that may be applicable.

[0040] Exemplary Methods

[0041] Figure 4 The figure is a schematic flow chart of a method for determining a base coordinate system provided by an exemplary embodiment of the present application. Figure 4 The method can be specifically executed by simulation software or electronic devices such as servers deployed with simulation software. For the convenience of description, the following description is given by taking the server executing the base coordinate system determination method as an example.

[0042] like Figure 4 As shown, the base coordinate system determination method may include the following contents.

[0043] 410: Determine a target three-dimensional model.

[0044] In some embodiments, after obtaining the three-dimensional model of the target object, the server can directly use the three-dimensional model of the target object as the target three-dimensional model, and adjust the target three-dimensional model according to a plurality of preset specific poses, and then perform collision simulation tests on the target three-dimensional model under each specific pose based on the three-dimensional plane in the simulation space. Finally, according to the pre-trained detection model, the target simulation result matching the target pose is detected from each simulation result, so as to construct the base coordinate system corresponding to the target simulation result according to the base coordinate system corresponding to the target pose. Therefore, the server can first determine the target three-dimensional model.

[0045] Specifically, the server may receive a three-dimensional model file sent by other devices, wherein the three-dimensional model file is used to store a three-dimensional model of a target object.

[0046] Next, the server can import the 3D model file into the simulation software, and in the simulation space corresponding to the simulation software, based on the correspondence between the components stored in the 3D model file and their base coordinate system, perform 3D reconstruction of the 3D model of the target object in the simulation space.

[0047] Finally, the server may use the three-dimensionally reconstructed three-dimensional model as the target three-dimensional model.

[0048] In one example, the 3D model file can be an OBJ file, a PLY file, or any file used to store a 3D model of a target object. The specific file type corresponding to the 3D model file can be set as needed, and this specification does not limit this.

[0049] In one example, the server has pre-stored three-dimensional model files, and when receiving an import instruction, the server determines the three-dimensional model file from the files stored in the server.

[0050] 420: Adjust the position and posture of the target three-dimensional model according to the multiple specific postures to obtain the target three-dimensional models corresponding to the multiple specific postures.

[0051] In some embodiments, the server stores a plurality of specific poses, and thus the server can adjust the pose of the target three-dimensional model for each specific pose to obtain the target three-dimensional models corresponding to the plurality of specific poses.

[0052] In one example, the server may determine a base coordinate system of the target three-dimensional model according to the three-dimensional model file, and adjust the position and posture of the target three-dimensional model according to each specific position and the base coordinate system.

[0053] In one example, the server may initialize the pose of the target 3D model, that is, consider the pose of the target 3D model as a default value. Then, the center of the target 3D model is determined, and the center of the target 3D model is used as a reference, and for each specific pose, the pose of the target 3D model is adjusted according to the specific pose.

[0054] 430: According to the three-dimensional plane in the simulation space, the target three-dimensional models corresponding to the multiple specific postures are simulated respectively to obtain the simulation results corresponding to the multiple specific postures. For each specific posture, the simulation result of the specific posture includes the target three-dimensional model of the specific posture being in a static posture after falling on the three-dimensional plane.

[0055] In some embodiments, the server may construct a three-dimensional plane in the simulation space of the simulation software, wherein, in the three-dimensional coordinate system corresponding to the simulation space, the coordinate values ​​of any two positions in the three-dimensional plane in the vertical axis of the three-dimensional coordinate system, that is, in the Z-axis direction, are the same value.

[0056] Afterwards, the server can configure appropriate environmental parameters for each 3D model. The environmental parameters can include gravity parameters and friction parameters. The gravity parameter can be 9.8 m / s 2 The friction parameter can be the friction coefficient, and the specific value can be 0.3, 0.5, etc.

[0057] Therefore, the server can simulate the target three-dimensional model corresponding to each specific posture according to the environmental parameters, so as to simulate the falling process in the simulation space and obtain the simulation result of the specific posture.

[0058] Among them, for each specific posture, the simulation result corresponding to the specific posture shows the static posture of the target three-dimensional model corresponding to the specific posture after it falls on the three-dimensional plane. The simulation result of the specific posture is the target three-dimensional model in a static state after the target three-dimensional model of the specific posture falls on the three-dimensional plane. In other words, the simulation result is actually the target three-dimensional model.

[0059] 440: Determine, from the simulation results, a simulation result that matches the target static posture by detecting the model as the target simulation result, and determine a base coordinate system of the target simulation result according to the base coordinate system that matches the target static posture.

[0060] In some embodiments, a trained detection model is stored in the server, and the detection model is used to determine whether each simulation result matches the target static posture.

[0061] Therefore, the server can determine the image data corresponding to each simulation result, and input the image data corresponding to the simulation result into the detection model to obtain the detection result output by the detection model. The detection result is used to indicate whether the image data input into the detection model matches the static posture of the target.

[0062] The server can then determine, based on the detection results output by the detection model, from the simulation results, a simulation result that matches the target static posture as the target simulation result.

[0063] Finally, the server may use the target simulation result as the target simulation result, and use the base coordinate system that matches the static posture of the target as the base coordinate system of the target simulation result.

[0064] In one example, if the target static posture is that the target object is placed upright on the three-dimensional plane, the base coordinate system corresponding to the target static posture may be a Cartesian coordinate system with the direction perpendicular to the three-dimensional plane as the vertical axis, that is, the Z axis.

[0065] In one example, if the target static posture is that the target object is placed upright on a three-dimensional plane and the orientation of the target object in the image data is frontal, then the base coordinate system corresponding to the target static posture can be a Cartesian coordinate system with the direction perpendicular to the three-dimensional plane as the Z axis, the projection of the center of the target three-dimensional model on the three-dimensional plane, and the direction of the projection of the image data acquisition device on the three-dimensional plane as the X axis.

[0066] A method for determining a base coordinate system is provided in an embodiment of the present application. After a target three-dimensional model is imported into simulation software, target three-dimensional models corresponding to multiple specific poses are obtained based on multiple preset specific poses and the base coordinate system of the target three-dimensional model. Then, according to the three-dimensional plane in the simulation space, the target three-dimensional models corresponding to the multiple specific poses are simulated to obtain multiple simulation results of the target three-dimensional model that may include the target three-dimensional model in a static pose after falling on the three-dimensional plane. Finally, by detecting the model, a target simulation result matching the static pose of the target is determined from each simulation result, and then the base coordinate system of the target simulation result is determined according to the base coordinate system matching the static pose of the target.

[0067] In this way, the embodiment of the present application can determine the target simulation result that matches the target static posture and the base coordinate system of the target simulation result from the target three-dimensional model corresponding to each specific posture and each simulation result of the three-dimensional plane in the simulation space through the detection model. Without the need for manual calibration of the base coordinate system, the base coordinate system that matches the target simulation result can be determined, thereby improving work efficiency, and is particularly suitable for scenarios where large-scale three-dimensional model data processing is required.

[0068] According to one embodiment of the present application, the base coordinate system determination method may also include: determining a specified number of first attitude angles and a specified number of second attitude angles, the first attitude angles and the second attitude angles corresponding to different attitude angle types, the first attitude angle is one of a roll angle, a heading angle and a pitch angle, and the second attitude angle is at least one of a roll angle, a heading angle and a pitch angle; combining the specified number of first attitude angles and the specified number of second attitude angles respectively to obtain a plurality of preset attitudes; and obtaining a plurality of specific attitudes based on the plurality of preset attitudes.

[0069] In one example, the specified number may be ten.

[0070] In an example, the first attitude angle may be one of a roll angle, a yaw angle, and a pitch angle, and the second attitude angle may be one of a roll angle, a yaw angle, and a pitch angle.

[0071] In one example, the server may store a plurality of heading angles, a plurality of pitch angles, and a plurality of roll angles, and the server may determine a specified number of first attitude angles and a specified number of second attitude angles from the stored plurality of heading angles, a plurality of pitch angles, and a plurality of roll angles according to the specified number, wherein the first attitude angle type corresponding to the first attitude angle is different from the attitude angle type corresponding to the second attitude angle.

[0072] In an example, the first attitude angle may be one of a roll angle, a yaw angle, and a pitch angle, and the second attitude angle may be two of a roll angle, a yaw angle, and a pitch angle.

[0073] In one example, the server may store a plurality of heading angles, a plurality of pitch angles, and a plurality of roll angles. Then, the server may determine a specified number of first attitude angles and a specified number of second attitude angles from the stored plurality of heading angles, a plurality of pitch angles, and a plurality of roll angles according to the specified number. The specified number of second attitude angles includes two different types of attitude angles. Taking the heading as the first attitude angle as an example, the specified number of second attitude angles may include a first number of pitch angles and a second number of roll angles, and the sum of the first number and the second number is the specified number.

[0074] In one example, the server may further specify a third attitude angle according to the specified number, and respectively combine the third attitude angle of the specified number and the first attitude angle of the specified number and the second attitude angle of the specified number to obtain a plurality of preset attitudes. The third attitude angle may be one of a roll angle, a heading angle, and a pitch angle, and the attitude angle type of the third attitude angle is different from the first attitude angle and / or the second attitude angle.

[0075] In one example, the server may determine a specified number of roll angles, a specified number of heading angles, and a specified number of pitch angles according to the specified quantity; combine the specified number of roll angles, the specified number of heading angles, and the specified number of pitch angles to obtain multiple preset postures; and obtain multiple specific postures based on the multiple preset postures.

[0076] In one example, the server may store multiple heading angles, multiple pitch angles, and multiple roll angles. The server may then determine a specified number of roll angles, a specified number of heading angles, and a specified number of pitch angles from the stored multiple heading angles, multiple pitch angles, and multiple roll angles based on the specified number.

[0077] In one example, for the value ranges corresponding to the roll angle, heading angle, and pitch angle, the server can sample them respectively to obtain ten different roll angles, ten different heading angles, and ten different pitch angles. Afterwards, the server can combine the collected ten different roll angles, ten different heading angles, and ten different pitch angles, respectively, so that 1000 preset postures can be obtained.

[0078] In one example, the sampling may be equally spaced sampling. For example, the dimension corresponding to the roll angle, that is, the value range corresponding to the roll angle, is 360 degrees in total. Then, performing 10 interval samplings will generate ten angles with an interval of 36 degrees from 0 to 360 degrees.

[0079] In one example, since the simulation in this application is a collision simulation under the action of gravity, the server may not change the plane position of the target three-dimensional model, and set the height of the target three-dimensional model to a specified height. Thus, the server may obtain multiple specific postures.

[0080] In one example, for target three-dimensional models of different sizes, the server may preset different designated quantities and different sampling ranges, that is, the value ranges of the roll angle, heading angle, and pitch angle.

[0081] In an example, the numbers of the sampled roll angles, heading angles, and pitch angles may be the same number or different numbers.

[0082] In this embodiment, by reasonably determining the posture sampling strategy, various possible posture spaces can be covered more accurately, so that when the posture of the target three-dimensional model imported into the simulation space is significantly different from the ideal state (taking the target object as a cup, for example, the cup mouth of the imported target three-dimensional model may be facing downward), the posture of the target three-dimensional model can be adjusted based on the sampled specific posture to ensure that among the adjusted target three-dimensional models, there is at least one target three-dimensional model that is less different from the ideal state, so as to ensure that subsequent simulation analysis can be carried out based on a more reasonable and accurate model posture, thereby improving the accuracy of the base coordinate system correction.

[0083] According to one embodiment of the present application, step 430 may include: for each specific posture, configuring environmental parameters for the target three-dimensional model corresponding to the specific posture, the environmental parameters including gravity parameters, friction parameters and elasticity parameters; determining the result after the target three-dimensional model corresponding to the specific posture collides with the three-dimensional plane in the simulation space under the environmental parameters, as the simulation result.

[0084] In one example, after determining the target three-dimensional model, the server may set the target three-dimensional model as a rigid body to ensure that it can correctly respond to physical effects during the simulation process.

[0085] In one example, the server may set the collision volume to SDF (signed distance function) or triangular patch to provide more sophisticated collision detection and make the physical simulation more realistic.

[0086] In one example, the three-dimensional plane in the simulation space may be a rigid plane, specifically a passive rigid body, and the length and width of the three-dimensional plane may be set to about n times its diameter according to the size of the target three-dimensional model, where n may be set as needed.

[0087] In an example, the environmental parameter may include an elastic parameter, which may also be called an elastic coefficient. The elastic coefficient may be set to 0 to prevent the target three-dimensional model from bouncing after colliding with the three-dimensional plane.

[0088] In this embodiment, by carefully setting the environmental parameters in the simulation engine, such as selecting a suitable collision volume (SDF), accurate gravity, friction and elastic coefficient, etc., a more realistic physical simulation can be achieved, so that a more accurate base coordinate system can be obtained based on the simulation results of the more realistic physical simulation.

[0089] According to one embodiment of the present application, step 440 may include: for each simulation result, determining the image data of the simulation result; inputting the image data into the detection model to obtain the detection result of the simulation result output by the detection model, and the detection result is used to characterize whether the image data matches the target static posture; and determining the simulation result corresponding to the target static posture according to the detection results corresponding to each simulation result.

[0090] Specifically, after the simulation is finished, for each simulation result, the server may determine the image data corresponding to the simulation result.

[0091] The server can then input the image data corresponding to each simulation result into the detection model respectively, and obtain the detection results corresponding to each simulation result output by the detection model.

[0092] Finally, the server may determine a simulation result that matches the static posture of the target from among the simulation results according to the detection results corresponding to the simulation results.

[0093] In one example, the detection model is a model trained based on methods in the field of artificial intelligence such as deep learning or reinforcement learning.

[0094] In one example, the detection result output by the detection model may be the probability that the simulation result matches the target posture.

[0095] In one example, the detection model is a pre-trained language model. Therefore, the server can determine the prompt text corresponding to the static posture of the target as the designated prompt text. And for each simulation result, the simulation result and the designated prompt text are combined and input into the detection model to obtain the detection result output by the detection model.

[0096] In one example, the specified prompt text is “determine whether the object in the image is perpendicular to the ground plane”, and the detection result output by the detection model is “yes” or “no”.

[0097] Based on this embodiment, a large number of simulation results are analyzed through the trained detection model, and simulation results that match the static posture of the target can be screened out with high accuracy. When processing complex three-dimensional models, the subjectivity and errors that may be caused by manual observation and judgment are avoided, and the accuracy of posture screening is greatly improved.

[0098] According to an embodiment of the present application, the detection model is a pre-trained language model, and step 440 may include: according to at least one acquisition point corresponding to the image acquisition device, for each simulation result, determining at least one image data corresponding to the simulation result; for each image data in the at least one image data, inputting the image data and the first prompt text into the detection model to obtain a first detection result of the image data output by the detection model, and the first detection result is used to characterize whether the simulation result is perpendicular to the three-dimensional plane at the acquisition perspective corresponding to the image data; when it is determined that the simulation result is perpendicular to the three-dimensional plane according to the first detection result corresponding to each image data in the at least one image data, the second detection result corresponding to each image data in the at least one image data is determined by the detection model and the second prompt text, and the second detection result is used to characterize whether the simulation result and the target static posture match at the acquisition perspective corresponding to the image data; when it is determined that the simulation result matches the target static posture according to the second detection result corresponding to each image data in the at least one image data, the simulation result is used as the target simulation result.

[0099] In one example, the step of determining the target simulation result from the simulation results can be divided into two stages. In the first stage, the detection model determines the simulation result perpendicular to the three-dimensional plane from each simulation result. In the second stage, the detection model determines the simulation result matching the target static posture from the simulation results perpendicular to the three-dimensional plane.

[0100] In one example, taking the target three-dimensional model as a car model, after the simulation is completed, the server can first determine the image data for each simulation result, which clearly shows the appearance of the car in different postures (such as the body at different angles, wheel positions, etc.). After that, the image data and the first prompt text (such as "determine whether the car in the image is perpendicular to the ground plane") are input into the pre-trained language model that has been trained to obtain the first detection result output by the pre-trained language model.

[0101] In one example, after receiving these inputs, the pre-trained language model uses its multimodal understanding of text and images to begin analyzing the relationship between the car posture in the image and the ground plane. It not only recognizes the pixel information in the image, but also combines the semantics of the prompt text to understand the concept of "vertical" and its manifestation in the car posture (such as the angle between the car body and the ground is close to 90 degrees, the contact between the wheels and the ground, etc.). After a complex internal reasoning process, the model outputs the first detection result of the simulation result, using a clear "yes" or "no" to characterize whether the simulation result is perpendicular to the three-dimensional plane. For example, if the model determines that the car body in the image is basically perpendicular to the ground and the wheels are in normal contact with the ground, it will output "yes"; otherwise, it will output "no".

[0102] In one example, when it is determined based on the first detection result that the simulation result is perpendicular to the three-dimensional plane (i.e., the model outputs "yes"), the server may input the image data and the second prompt text (such as "determine whether the posture of the car in the image matches the target posture, the target posture is that the front of the car is facing straight ahead, the body is horizontal, and the wheels are placed vertically on the ground") into the pre-trained language model again. At this time, the model will further analyze the details of the car's posture in the image, such as whether the direction of the front of the car is accurate, whether the body is completely horizontal, whether the wheels are perpendicular to the ground, and whether it meets the target static posture requirements. After more sophisticated reasoning, the model outputs the second detection result of the simulation result, which also uses "yes" or "no" to characterize whether it matches the target static posture. For example, if the front of the car deviates slightly from the front, the model may output "no"; only when the car posture fully meets the requirements of the target driving posture, the model will output "yes".

[0103] In one example, when it is determined according to the second detection result that the simulation result matches the target static posture (i.e., the model output is "yes"), the server may use the simulation result as the target simulation result. For the car model, this target simulation result is the simulation result that best matches the expected car posture (the front of the car is facing straight ahead, the body is level, etc.), which will serve as the key model state for subsequent applications such as car autonomous driving simulation and car performance testing (such as wind resistance testing, which requires accurate posture).

[0104] In one example, for each simulation result, the server can collect image data from multiple angles through image acquisition devices set at different acquisition locations, just like shooting an actor's performance from multiple audience perspectives. Each simulation result will correspond to multiple image data (such as images taken from the front, side, top, etc.). Therefore, based on at least one acquisition location corresponding to the image acquisition device, the server can determine at least one image data corresponding to the simulation result.

[0105] In one example, the server may input each of the image data and the first prompt text into a pre-trained language model, that is, a detection model, in sequence, to obtain first detection results corresponding to each image data output by the detection model.

[0106] In one example, when it is determined that the simulation result is perpendicular to the three-dimensional plane according to the first detection results corresponding to each image data, the server can determine the second detection results corresponding to each image data through the detection model and the second prompt text, and when it is determined that the simulation result matches the target static posture according to the second detection results corresponding to each image data, the simulation result is used as the target simulation result.

[0107] In this embodiment, the multimodal understanding ability of the pre-trained language model is used to combine image data and prompt text for posture judgment, which can more comprehensively and accurately understand the posture characteristics of the model and improve the reliability of the judgment result. At the same time, through the staged prompt text input and judgment, it can be flexibly adjusted according to different target static posture requirements, which enhances the flexibility and adaptability of this method in automotive applications.

[0108] According to an embodiment of the present application, step 440 may include: when a first detection result corresponding to each image data in at least one image data is perpendicular to the three-dimensional plane, determining that the simulation result is perpendicular to the three-dimensional plane.

[0109] Specifically, taking the target three-dimensional model as a character model as an example, for each simulation result, the server can use the image acquisition devices at these different acquisition locations to collect image data from multiple angles, just like shooting an actor's performance from multiple audience perspectives. Each simulation result will correspond to multiple image data (such as images taken from the front, side, top, etc.).

[0110] Then, each of these image data and the first prompt text (such as "determine whether the person in the image is perpendicular to the ground plane") are sequentially input into the pre-trained language model, that is, the detection model. The model analyzes each image data and outputs a first detection result ("yes" or "no").

[0111] Finally, when the first detection results of all image data are yes, or in other words, the first detection results of all images can be used to indicate that the simulation result is perpendicular to the three-dimensional plane (that is, all images show that the person is standing basically vertically on the ground), the simulation result is determined to be perpendicular to the three-dimensional plane. This step is like multiple judges judging the actor's posture from different angles. Only when all judges believe that the actor's posture is vertical, it is determined to be vertical.

[0112] Based on this embodiment, the overall misjudgment caused by partial perspective misjudgment can be effectively avoided. It is ensured that only when the simulation result is truly perpendicular to the plane from all possible observation angles, it is considered vertical, which provides a solid foundation for subsequent more accurate posture matching judgment and improves the accuracy and reliability of the entire posture judgment process.

[0113] According to one embodiment of the present application, the base coordinate system determination method may also include: determining a hemisphere with a center of the simulation result as the center and a specified distance as the radius, and deploying at least one acquisition point of the image acquisition device on the hemisphere, wherein each acquisition point in the at least one acquisition point corresponds to a non-completely identical azimuth angle.

[0114] In one example, the acquisition point of the image acquisition device during acquisition can be sampled in the following manner: a hemispherical surface is constructed with the center of the target three-dimensional model as the center and a specified distance as the radius. After the hemispherical surface is constructed, the server can sample 10 horizontal angles on the hemispherical surface according to the horizontal plane of the sphere, and set three pitch angles of 0 degrees, 20 degrees, and 40 degrees in sequence, thereby obtaining 30 sampling points. Finally, according to the formula x=Rcos(θ)cos(φ), y=Rcos(θ)sin(φ), z=Rsin(θ), the position of the image acquisition device is calculated (where θ is the pitch angle and φ is the horizontal angle) as the acquisition point of the image acquisition device.

[0115] In one example, the server may also randomly sample on the hemispherical surface and use the sampling positions as the acquisition points of the image acquisition device.

[0116] In one example, each acquisition point corresponds to a different azimuth, and the straight line formed by the projection of the acquisition point corresponding to the same azimuth on the three-dimensional plane and the projection of the center of the hemisphere on the three-dimensional plane is the same straight line, such as Figure 5 shown. Figure 5 A schematic diagram of the acquisition points corresponding to the same azimuth provided in this specification. In which, point O is the center point of the hemisphere, the black rectangles are three different acquisition points, and the plane where point O is located is the three-dimensional plane constructed in step 420, or the plane where point O is located is parallel to the three-dimensional plane constructed in step 420. For example, it is obvious that for each acquisition point, the straight line formed by the projection of the acquisition point on the three-dimensional plane and the projection of the center of the hemisphere on the three-dimensional plane is a straight line formed by the projection of any one of the remaining acquisition points in the figure on the three-dimensional plane and the projection of the center of the hemisphere on the three-dimensional plane.

[0117] It should be noted that the azimuth corresponding to the acquisition point is used to characterize the relative position between the projection point of the acquisition point in the three-dimensional plane and the projection point of the center of the hemisphere on the three-dimensional plane. It only considers the directional relationship between the projection points of the two in the three-dimensional plane, and has nothing to do with the difference between the acquisition point and the center of the hemisphere in the direction perpendicular to the three-dimensional plane.

[0118] In this way, the multi-angle image data provided by multiple acquisition points can complement and verify each other, which can improve the accuracy and reliability of posture judgment.

[0119] According to an embodiment of the present application, step 440 may include: dividing the image data corresponding to the simulation result into multiple groups of image data according to the acquisition points corresponding to each image data in at least one image data, and for each group of image data, the acquisition points of each image data contained in the group of image data correspond to the same azimuth; for each group of image data, determining the number of image data in the group of image data whose second detection results match the simulation result and the target static posture; if the number is greater than a number threshold, determining that the simulation result at the azimuth corresponding to the group of image data is the target simulation result.

[0120] In one example, when capturing images from different azimuth angles, the viewing angle of the image capture device relative to the target object changes. Therefore, what is captured at the capture points at different azimuth angles is actually a target 3D model with different postures. When determining whether the target object corresponding to the target 3D model is perpendicular to the desktop, the image data captured at different capture points can be used. When determining the front of the target 3D model, it is inevitable to consider the influence of the azimuth angle of the image capture device.

[0121] In one example, the server may divide the image data corresponding to the simulation result into multiple groups of image data according to the azimuth angles corresponding to the respective acquisition points. For each group of image data, the acquisition points of the image data contained in the group of image data correspond to the same azimuth angles.

[0122] In one example, if there are five images in a group of image data, and these five images include five images taken from the same azimuth but from different pitch angles, if the second detection results of four of these five images are "yes", when the quantity threshold is set to 3 (which can be adjusted according to actual needs and accuracy requirements), since 4 is greater than 3, it can be determined that the posture of the simulation result at the azimuth corresponding to the group of images matches the target static posture. Therefore, the server can determine that the simulation result at the azimuth corresponding to the group of image data is the target simulation result.

[0123] In one example, the server may select a specific number of image data from each image data corresponding to the simulation result for each simulation result as image data for determining the first detection result, and then determine that the simulation result is perpendicular to the three-dimensional plane when the first detection results of all image data used to determine the first detection result are that the simulation result is perpendicular to the three-dimensional plane. The number of image data used to determine the first detection result, that is, the specific number, is less than the number of all image data corresponding to the simulation result.

[0124] In one example, the server may determine a specific number of acquisition points from each acquisition point according to the specific number as specific points, and use the image data acquired by the image acquisition device at the specific points as image data for determining the first detection result. Taking 30 acquisition points as an example, the number of the specific points may be 3, 4, 5, or other integers less than 30.

[0125] In one example, after determining the simulation result perpendicular to the three-dimensional plane, the server can determine each image data corresponding to the simulation result. Taking the image data as the image data collected by the image acquisition device at the aforementioned 30 points as an example, the 30 images can be divided into 10 groups according to the azimuth sampling, with 3 overhead angles in each group. Therefore, for each group of images, if there are 2 or more images in the group that are judged to be the front of the object, then the azimuth corresponding to the group of images is judged to be perpendicular or approximately perpendicular to the front of the object.

[0126] According to one embodiment of the present application, step 440 may include: determining a direction perpendicular to the three-dimensional plane as the vertical axis of the base coordinate system of the target simulation result; determining the direction of the azimuth corresponding to the target simulation result as the horizontal axis of the base coordinate system of the target simulation result; and constructing the base coordinate system of the target simulation result based on the horizontal axis and the vertical axis.

[0127] Specifically, since the target simulation result is standing upright on the desktop, the server can determine the direction perpendicular to the three-dimensional plane in step 420 as the vertical axis of the base coordinate system of the target simulation result, that is, the Z axis. When the target simulation result is determined, that is, when the front of the target object is determined, since the azimuth angle of the image acquisition device is usually perpendicular to the front of the target object, the server can determine the direction of the azimuth angle corresponding to the target simulation result as the direction of the horizontal axis of the base coordinate system of the target simulation result. Finally, based on the right-hand rule, the server can construct the base coordinate system of the target simulation result.

[0128] In this way, after completing the natural posture judgment and the front posture judgment of the object, the judgment of the rotation amount of the object base coordinate system can be completed.

[0129] Figure 6 Shown is a schematic flow chart of a method for determining a base coordinate system provided by another exemplary embodiment of the present application. Figure 6 The embodiment is Figure 4 For the example of the embodiment, in order to avoid repetition, the same points can be referred to the description in the above embodiment, which will not be repeated here. Figure 6 As shown, the base coordinate system determination method may include the following contents.

[0130] 610: Determine a target three-dimensional model.

[0131] 620: Obtain target three-dimensional models corresponding to multiple specific poses.

[0132] 660: Simulate the target three-dimensional models corresponding to the multiple specific postures respectively to obtain simulation results.

[0133] 640: For each simulation result, determine a first detection result of at least one image data corresponding to the simulation result by using the detection model and the first prompt text.

[0134] 650: Split at least one image data into multiple groups of image data according to the acquisition position of each image data, and when it is determined according to the first detection result that the simulation result is perpendicular to the three-dimensional plane, for each group of image data, determine the second detection result of each image data in the group of image data through the detection model and the second prompt text.

[0135] 660: When it is determined that the group of image data matches the target static posture according to the second detection result of each image data in the group of image data, a simulation result at the azimuth angle corresponding to the group of image data is determined as the target simulation result.

[0136] 670: Construct a base coordinate system of the target simulation result based on the azimuth angle of the set of image data.

[0137] The execution order of the above steps 610-670 can be adjusted as needed. And the above steps 610-670 can refer to the description of steps 410-440 in this specification, which will not be repeated in this specification. The base coordinate system determination method provided in the embodiment of the present application can realize the determination of the base coordinate system of the target simulation result.

[0138] Exemplary Devices

[0139] Figure 7 FIG. 7 is a schematic diagram of a base coordinate system determining device 700 provided by an exemplary embodiment of the present application. Figure 7As shown, the base coordinate system determination device 700 includes: an acquisition module 710, which is used to determine a target three-dimensional model; an adjustment module 720, which is used to adjust the position and posture of the target three-dimensional model according to a plurality of specific postures, and obtain target three-dimensional models corresponding to the plurality of specific postures; a simulation module 730, which is used to simulate the target three-dimensional models corresponding to the plurality of specific postures according to a three-dimensional plane in a simulation space, and obtain simulation results corresponding to the plurality of specific postures, wherein for each specific posture, the simulation result of the specific posture includes a target three-dimensional model of the target three-dimensional model in the specific posture that is in a static posture after falling on the three-dimensional plane; a determination module 740, which is used to determine, from each simulation result, a simulation result that matches the static posture of the target as the target simulation result by detecting a model, and determine the base coordinate system of the target simulation result according to the base coordinate system that matches the static posture of the target.

[0140] According to one embodiment of the present application, the simulation module 730 is used to: determine a specified number of first attitude angles and a specified number of second attitude angles, the first attitude angles and the second attitude angles correspond to different attitude angle types, the first attitude angle is one of a roll angle, a heading angle and a pitch angle, and the second attitude angle is at least one of a roll angle, a heading angle and a pitch angle; respectively combine the specified number of first attitude angles and the specified number of second attitude angles to obtain a plurality of preset attitudes; and obtain a plurality of specific attitudes based on the plurality of preset attitudes.

[0141] According to one embodiment of the present application, the simulation module 730 is used to: for each specific posture, configure environmental parameters for the target three-dimensional model corresponding to the specific posture, the environmental parameters including gravity parameters, friction parameters and elasticity parameters; determine the result of the target three-dimensional model corresponding to the specific posture colliding with the three-dimensional plane in the simulation space under the environmental parameters as the simulation result.

[0142] According to one embodiment of the present application, the determination module 740 is used to: determine the image data of each simulation result; input the image data into the detection model to obtain the detection result of the simulation result output by the detection model, and the detection result is used to characterize whether the image data matches the target static posture; determine the simulation result corresponding to the target static posture according to the detection results corresponding to each simulation result.

[0143] According to an embodiment of the present application, the determination module 740 is used to: determine, for each simulation result, at least one image data corresponding to the simulation result according to at least one acquisition point corresponding to the image acquisition device; for each image data in the at least one image data, input the image data and the first prompt text into the detection model to obtain a first detection result of the image data output by the detection model, and the first detection result is used to characterize whether the simulation result is perpendicular to the three-dimensional plane at the acquisition viewing angle corresponding to the image data; when it is determined that the simulation result is perpendicular to the three-dimensional plane according to the first detection result corresponding to each image data in the at least one image data, the second detection result corresponding to each image data in the at least one image data is determined by the detection model and the second prompt text, and the second detection result is used to characterize whether the simulation result and the target static posture match at the acquisition viewing angle corresponding to the image data; when it is determined that the simulation result matches the target static posture according to the second detection result corresponding to each image data in the at least one image data, the simulation result is used as the target simulation result.

[0144] According to an embodiment of the present application, the determination module 740 is used to: determine that the simulation result is perpendicular to the three-dimensional plane when the first detection result corresponding to each image data in the at least one image data is perpendicular to the three-dimensional plane.

[0145] According to one embodiment of the present application, the determination module 740 is used to: determine a hemisphere with a center of the simulation result as the center and a specified distance as the radius, and deploy at least one acquisition point of the image acquisition device on the hemisphere, wherein each acquisition point in the at least one acquisition point corresponds to a non-completely identical azimuth angle.

[0146] According to an embodiment of the present application, a determination module 740 is used to: divide the image data corresponding to the simulation result into multiple groups of image data according to the acquisition points corresponding to each image data in at least one image data, and for each group of image data, the acquisition points of each image data contained in the group of image data correspond to the same azimuth; for each group of image data, determine the number of image data in the group of image data whose second detection results match the simulation result and the target static posture; if the number is greater than a number threshold, determine that the simulation result at the azimuth corresponding to the group of image data is the target simulation result.

[0147] According to one embodiment of the present application, the determination module 740 is used to: determine the direction perpendicular to the three-dimensional plane as the vertical axis of the base coordinate system of the target simulation result; determine the direction of the azimuth corresponding to the target simulation result as the horizontal axis of the base coordinate system of the target simulation result; and construct the base coordinate system of the target simulation result based on the horizontal axis and the vertical axis.

[0148] An embodiment of the present application also provides a robot, which includes a control module, and the control module is used to execute the base coordinate system determination method provided in any of the above embodiments.

[0149] The operation and functions of the robot provided in the embodiment of the present application can refer to the above Figure 4 or Figure 6 The description of the base coordinate system determination method provided in any of the embodiments will not be repeated here to avoid repetition.

[0150] Figure 8 The block diagram of an electronic device 800 for executing a method for determining a base coordinate system provided by an exemplary embodiment of the present application is shown. The electronic device 800 may be a mobile device, a control device of a mobile device, a server or other device interacting with a mobile device.

[0151] Reference Figure 8 , the electronic device 800 includes a processing component 810, which further includes one or more processors, and a memory resource represented by a memory 820 for storing instructions executable by the processing component 810, such as an application. The application stored in the memory 820 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 810 is configured to execute instructions to perform the above-mentioned base coordinate system determination method.

[0152] The electronic device 800 may also include a power supply component configured to perform power management of the electronic device 800, a wired or wireless network interface configured to connect the electronic device 800 to a network, and an input / output (I / O) interface. The electronic device 800 may be operated based on an operating system stored in the memory 820, such as Windows Server 2000. TM , Mac OSX TM , Unix TM , Linux TM , FreeBSD TM or similar.

[0153] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of the electronic device 800, enables the electronic device 800 to execute a base coordinate system determination method.

[0154] An embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor of a computer device, the computer device can execute the base coordinate system determination method provided in any of the above embodiments.

[0155] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, and will not be described one by one here.

[0156] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can 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 this application.

[0157] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0158] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0159] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0160] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0161] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program check codes.

[0162] It should be noted that, in the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and cannot be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" is two or more.

[0163] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0164] The above are only preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for determining a base coordinate system, characterized in that: include: Determine the target three-dimensional model; According to the multiple specific poses, the pose of the target three-dimensional model is adjusted to obtain the target three-dimensional models corresponding to the multiple specific poses respectively; According to the three-dimensional plane in the simulation space, the target three-dimensional models corresponding to the multiple specific postures are simulated respectively to obtain simulation results corresponding to the multiple specific postures, and for each specific posture, the simulation result of the specific posture includes a target three-dimensional model in a static posture after the target three-dimensional model in the specific posture falls on the three-dimensional plane; By detecting the model, a simulation result matching the target static posture is determined from each simulation result as the target simulation result, and a base coordinate system of the target simulation result is determined according to the base coordinate system matching the target static posture.

2. The method for determining a base coordinate system according to claim 1, characterized in that: Also includes: Determining, according to a specified number, a first attitude angle of the specified number and a second attitude angle of the specified number, wherein the first attitude angle and the second attitude angle correspond to different attitude angle types, the first attitude angle is one of a roll angle, a heading angle, and a pitch angle, and the second attitude angle is at least one of a roll angle, a heading angle, and a pitch angle; Combining the specified number of first posture angles and the specified number of second posture angles respectively to obtain a plurality of preset postures; The multiple specific postures are obtained according to the multiple preset postures.

3. The method for determining a base coordinate system according to claim 1, wherein: The method of simulating the target three-dimensional models respectively corresponding to the multiple specific postures according to the three-dimensional plane in the simulation space to obtain the simulation results respectively corresponding to the multiple specific postures includes: For each specific posture, configure environmental parameters for the target three-dimensional model corresponding to the specific posture, wherein the environmental parameters include gravity parameters, friction parameters and elastic parameters; The result after the target three-dimensional model corresponding to the specific posture collides with the three-dimensional plane in the simulation space under the environmental parameters is determined as a simulation result.

4. The method for determining a base coordinate system according to claim 1, wherein: The method of determining the simulation result matching the static posture of the target from the simulation results by detecting the model includes: For each simulation result, determining image data of the simulation result; Inputting the image data into the detection model to obtain a detection result of the simulation result output by the detection model, wherein the detection result is used to indicate whether the image data matches the static posture of the target; According to the detection results corresponding to the simulation results, a simulation result matching the static posture of the target is determined.

5. The method for determining a base coordinate system according to claim 1, characterized in that: The detection model is a pre-trained language model; The method of determining, from the simulation results, a simulation result matching the target static posture as the target simulation result by detecting the model, includes: Determine, for each simulation result, at least one image data corresponding to the simulation result according to at least one acquisition point corresponding to the image acquisition device; For each image data in the at least one image data, the image data and the first prompt text are input into the detection model to obtain a first detection result of the image data output by the detection model, wherein the first detection result is used to indicate whether the simulation result is perpendicular to the three-dimensional plane at a collection viewing angle corresponding to the image data; When it is determined that the simulation result is perpendicular to the three-dimensional plane according to the first detection results respectively corresponding to each image data in the at least one image data, a second detection result respectively corresponding to each image data in the at least one image data is determined by the detection model and the second prompt text, wherein the second detection result is used to indicate whether the simulation result matches the static posture of the target at the acquisition viewing angle corresponding to the image data; When it is determined that the simulation result matches the target static posture according to the second detection results respectively corresponding to each image data in the at least one image data, the simulation result is used as the target simulation result.

6. The method for determining a base coordinate system according to claim 5, characterized in that: Determining, according to first detection results respectively corresponding to each image data in the at least one image data, that the simulation result is perpendicular to the three-dimensional plane includes: When the first detection result corresponding to each image data in the at least one image data is perpendicular to the three-dimensional plane, it is determined that the simulation result is perpendicular to the three-dimensional plane.

7. The method for determining a base coordinate system according to claim 5, characterized in that: Also includes: Determine a hemisphere with a center of the simulation result as the center and a specified distance as the radius, and deploy the at least one acquisition point of the image acquisition device on the hemisphere, wherein each acquisition point in the at least one acquisition point corresponds to a non-completely identical azimuth angle.

8. The method for determining a base coordinate system according to claim 7, characterized in that: When it is determined that the simulation result matches the target static posture according to the second detection results respectively corresponding to each image data in the at least one image data, taking the simulation result as the target simulation result, comprises: According to the acquisition points respectively corresponding to the respective image data in the at least one image data, the image data corresponding to the simulation result is divided into a plurality of groups of image data, wherein for each group of image data, the acquisition points of the respective image data included in the group of image data correspond to the same azimuth; For each set of image data, determining the number of image data in the set of image data whose second detection result is that the simulation result matches the static posture of the target; If the number is greater than the number threshold, the simulation result at the azimuth angle corresponding to the group of image data is determined to be the target simulation result.

9. The method for determining a base coordinate system according to claim 8, characterized in that: The step of determining a base coordinate system of the target simulation result according to a base coordinate system matching the static posture of the target includes: Determine a direction perpendicular to the three-dimensional plane as the vertical axis of the base coordinate system of the target simulation result; Determine the direction of the azimuth angle corresponding to the target simulation result as the horizontal axis of the base coordinate system of the target simulation result; A base coordinate system of the target simulation result is constructed according to the horizontal axis and the vertical axis.

10. A robot, characterized in that: The robot comprises a control module, and the control module is used to execute the base coordinate system determination method according to any one of claims 1 to 9.

11. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor, Wherein, the processor is used to execute the base coordinate system determination method described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and the computer program is used to execute the base coordinate system determination method according to any one of claims 1 to 9.

13. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor of a computer device, the computer device is enabled to execute the base coordinate system determination method according to any one of claims 1 to 9.

Images