A three-dimensional space accessibility judgment and immersion entry generation method and system based on physical constraints of a subject and a storage medium

Abstract

A three-dimensional space accessibility judgment and immersion entry generation method based on physical constraints of a subject, a system and a storage medium, the method comprising: acquiring a three-dimensional space representation and physical constraint parameters of a subject to be entered; generating or acquiring space accessibility information, the information including support area, clearance area and passable area or boundary constraint; verifying, fusing or reconstructing the information according to the physical constraint parameters of the subject to obtain a subjectized space accessibility result; generating entry pose, actionable area and physical boundary constraint based on the result and outputting, so that the runtime environment controls the camera pose or the subject movement boundary accordingly. The present application can improve the matching degree of the entry pose and the subject geometric scale or the action constraint, reduce the camera entry obstacles, the subject out-of-bound and the invalid entry, and avoid the inadaptation of the existing space information to the specific subject.

Description

Technical Field

[0001] This invention relates to the fields of three-dimensional spatial computing, three-dimensional model runtime control, computer graphics, extended reality, and human-computer interaction. In particular, it relates to a method, system, and storage medium for determining the spatial accessibility of a three-dimensional spatial representation based on the physical constraint parameters of the subject, and generating the entrance pose, the movable area, and the physical boundary constraints.

[0002] This invention can be applied to mobile 3D spatial browsing, 3D spatial content preview, digital museums, digital cities, 3D spatial organization based on user history content, presentation and interactive environments, extended reality environments, 3D runtime engines, robot simulation, or other computer implementation environments that require importing subjects into 3D content and restricting their movement boundaries. Background Technology

[0003] With the development of 3D scanning, AI-generated content, extended reality, digital twins, and 3D modeling technologies, the sources of 3D content are becoming increasingly diversified. 3D content can be generated from laser scanning, structured indoor scanning, photogrammetry, modeling software, AI-generated systems, digital twin systems, extended reality scene understanding systems, or other 3D spatial data formats.

[0004] Current methods for displaying 3D content typically treat the input 3D model as an externally observable object. For example, the system can determine the position of the external camera based on the bounding box of the 3D model, bringing the entire model into the field of view. This method is suitable for displaying ordinary 3D objects such as furniture, artifacts, and sculptures. However, for accessible 3D spaces such as rooms, exhibition halls, streets, and museums, the external observation method cannot naturally place the subject inside the space. Users usually need to manually zoom, rotate, or move the viewpoint to attempt to enter the model's interior, and during this process, the camera may be located outside walls, obstacles, the ground, or other invalid areas.

[0005] Existing game scenarios, digital exhibition halls, or extended reality applications typically rely on pre-made scenes, manually configured entrances, pre-made colliders, navigation areas, or runtime movement boundaries. This approach is suitable for fixed 3D scenes that developers have pre-created and debugged, but it's difficult to directly apply to open-source 3D spaces. For open-source 3D space representations such as user-uploaded scan data, AI-generated 3D content, or unconfigured 3D models, the system often cannot pre-determine, as with pre-made scenes, where one can stand, where there is sufficient clearance, where movement is possible, or where obstacles are located. It also cannot determine whether existing candidate entrances, navigation areas, collision boundaries, or spatial semantics are suitable for the current subject.

[0006] Existing indoor navigation, path planning, accessible area assessment, or accessibility map generation technologies can extract paths, road networks, or accessible areas from 3D data, but their primary purpose is usually to generate offline navigation data, path planning results, access cost maps, or accessibility maps. These technologies typically do not directly address the runtime entry process of 3D content, nor do they use entrance poses, movable areas, and physical boundary constraints as a combined output to control runtime camera poses or subject movement boundaries.

[0007] Furthermore, even if existing technologies can provide candidate entrances, navigation areas, collision boundaries, ground, walls, or spatial semantic information, they typically lack mechanisms for verification, fusion, or reconstruction in conjunction with specific subject geometry or action constraints. For example, the same entrance or passable area may have different accessibility for ordinary adults, children, wheelchair users, digital subjects, or robotic subjects; the same collision boundary or navigation area may also produce different adaptation results due to differences in subject envelope size, turning radius, viewpoint height, or traversable height.

[0008] Therefore, a new computer implementation scheme is needed to enable the system to determine spatial accessibility based on open-source 3D spatial representations and combine them with the physical constraint parameters of the subject, and to generate entrance poses, movable areas and physical boundary constraints to control the camera poses or subject movement boundaries during runtime, thereby improving the effectiveness and stability of the 3D spatial entry process. Summary of the Invention

[0009] (a) Purpose of the invention This invention addresses at least the following technical problems: 1. In open-source 3D spatial representations, how to obtain spatial accessibility information that can be used for subject entry and action control; 2. How to incorporate the subject's geometric scale or action constraints into the 3D spatial entrance generation process; 3. How to verify whether existing candidate entry points, navigation areas, collision boundaries, or spatial semantics are suitable for the current subject; 4. How to fuse or reconstruct existing spatial accessibility information when it is incomplete, inconsistent, or does not meet the main physical constraint parameters; 5. How to generate entrance pose, movable area, and physical boundary constraints to control the camera pose or subject movement boundary during runtime.

[0010] (II) Technical Solution To address the aforementioned technical problems, this invention provides a method for determining the accessibility of a three-dimensional space and generating an immersive entrance based on the physical constraints of the subject. The method includes: Obtain a three-dimensional spatial representation, which is used to describe the three-dimensional content to be entered; Obtain the physical constraint parameters of the subject, which are used to characterize the geometric scale or action constraints of the subject to enter the three-dimensional content; Spatial calculations are performed on the three-dimensional spatial representation to generate spatial accessibility information; if the three-dimensional spatial representation or its associated data already contains spatial accessibility information, the spatial accessibility information is read or received; wherein, the spatial accessibility information is used to characterize the spatial conditions in the three-dimensional content that allow the subject to enter or move, and the spatial accessibility information includes at least one of a support area and a clear area, as well as a passable area or a boundary constraint; Based on the physical constraint parameters of the subject, the spatial accessibility information is verified, fused, or reconstructed to obtain a subjectified spatial accessibility result that matches the subject. Based on the subjectified space accessibility results, the entrance pose, movable area, and physical boundary constraints are generated. The entry pose, movable area, and physical boundary constraints are output so that the runtime environment controls the camera pose or the subject movement boundary according to the entry pose, movable area, and physical boundary constraints, and imports the subject into the 3D content.

[0011] The aforementioned branches that perform spatial calculations on the representation of three-dimensional space, as well as those that read or receive existing spatial accessibility information, all share a common technical concept: to verify, fuse, or reconstruct spatial accessibility information by combining the main physical constraint parameters, and to generate entrance poses, movable areas, and physical boundary constraints.

[0012] In this specification, the input / output pose, movable area, and physical boundary constraints may include passing the above data to the runtime module, writing it to the storage medium, transmitting it through the network interface, returning it through the application programming interface, or providing it to other system modules.

[0013] In some embodiments, verification, fusion, or reconstruction can be performed selectively or in combination based on the source, completeness, and credibility of the spatial accessibility information. When the existing spatial accessibility information is complete and matches the subject's physical constraint parameters, a subjectified spatial accessibility result can be obtained through verification; when the spatial accessibility information comes from multiple sources and there are differences between the sources, a subjectified spatial accessibility result can be obtained through fusion; when the spatial accessibility information is missing, incomplete, or does not match the subject's physical constraint parameters, a subjectified spatial accessibility result can be obtained through reconstruction.

[0014] In some embodiments, the subject is a human user or a digital subject corresponding to a human user; the entry pose is used to enable the human user to enter the three-dimensional content from a first-person perspective, and to match the first-person perspective with the physical constraint parameters of the subject.

[0015] In some embodiments, the physical constraint parameters of the subject include at least one of viewpoint height, subject envelope size, movement radius, turning radius, or traversable height.

[0016] In some embodiments, the spatial accessibility information includes a support area, a clearance area, a passable area, and boundary constraints; wherein the support area is used to determine the position where the subject can stand or stop, the clearance area is used to determine the collision-free space of the subject's envelope in the height or volume direction, the passable area is used to determine the range of movement of the subject, and the boundary constraints are used to restrict the subject from crossing obstacles or spatial boundaries.

[0017] In some embodiments, when the three-dimensional spatial representation or its associated data already contains entry-type information or boundary-type information, verification is performed based on the physical constraint parameters of the subject; wherein, the entry-type information includes candidate entry points or navigation areas, and the boundary-type information includes collision boundaries or spatial semantics; the verification includes verifying whether the entry-type information satisfies the subject entry conditions, and verifying whether the boundary-type information satisfies the subject movement boundary conditions.

[0018] In some embodiments, when existing spatial accessibility information is incomplete, inconsistent, or does not meet the subject physical constraint parameters, spatial accessibility information from multiple sources is fused, or at least a portion of the spatial accessibility information is reconstructed based on the three-dimensional spatial representation.

[0019] In some embodiments, the entrance pose includes an entrance position and an entrance orientation. The entrance position is set within an accessible area that matches the physical constraint parameters of the main body. The entrance orientation is preferentially determined based on the open space direction. When the open space direction does not meet preset conditions, it is determined based on the main axis of space or a candidate entrance. When there are task objectives or boundary constraints, the entrance orientation is corrected based on the task objectives or boundary constraints.

[0020] In some embodiments, the movable area and the physical boundary constraints are used to control the movement boundaries of the subject in the three-dimensional content, so that the subject is restricted to the area that can be entered in the runtime environment and avoids crossing at least one of obstacle boundaries, collision boundaries or ground boundaries.

[0021] This invention also provides a system for determining the accessibility of a three-dimensional space and generating an immersive entrance based on the physical constraints of the subject. The system includes: The spatial representation acquisition module is used to acquire a three-dimensional spatial representation; The main constraint acquisition module is used to acquire the physical constraint parameters of the main body; The spatial computing module is used to perform spatial calculations on the three-dimensional spatial representation to generate spatial accessibility information, or to read or receive the spatial accessibility information when the three-dimensional spatial representation or its associated data already contains spatial accessibility information. The subjectification processing module includes a verification unit and a fusion reconstruction unit. The verification unit is used to verify the spatial accessibility information according to the physical constraint parameters of the subject after the spatial calculation module reads or receives the spatial accessibility information. The fusion reconstruction unit is used to fuse spatial accessibility information from multiple sources or reconstruct at least a part of the spatial accessibility information based on the three-dimensional spatial representation when the verification fails, the spatial accessibility information is incomplete or inconsistent, so as to obtain a subjectified spatial accessibility result that matches the subject. The entrance generation module is used to generate the entrance pose, movable area and physical boundary constraints based on the subjectified space accessibility results. The runtime control module is used to control the camera pose or the subject movement boundary based on the entrance pose, the movable area, and the physical boundary constraints.

[0022] The present invention also provides a computer-readable storage medium storing a computer program or instructions that, when executed by a processor, implement the above-described method.

[0023] (III) Beneficial Effects Compared with the prior art, the present invention has at least the following beneficial effects: 1. This invention processes three-dimensional spatial representation and spatial accessibility information, and can transform the process of entering three-dimensional content from external bounding box framing to spatial accessibility calculation; 2. This invention introduces physical constraint parameters for the main body, enabling the entrance pose, movable area, and physical boundary constraints to match the geometric scale or movement constraints of the main body; 3. This invention can reduce the probability that the entrance pose is located in a wall, obstacle, suspended position, or other invalid area; 4. By constraining the movable area and physical boundaries, this invention can reduce the risk of the subject crossing obstacle boundaries, collision boundaries, or ground boundaries; 5. This invention can perform subject-based verification of existing candidate entrances, navigation areas, collision boundaries, or spatial semantics, avoiding the direct use of existing spatial information that is not compatible with the current subject; 6. This invention can fuse or reconstruct existing spatial accessibility information when it is incomplete, inconsistent, or fails verification, thereby improving the adaptability of open-source 3D spatial input; 7. The present invention is compatible with three-dimensional spatial representations from different sources, such as scanning space, modeling space, artificial intelligence generated space, extended reality scene understanding results, digital museum space, digital city space, and robot task space. Attached Figure Description

[0024] Figure 1 is a schematic diagram of the overall process of the method for determining the accessibility of a three-dimensional space and generating an immersive entrance based on the physical constraints of the subject provided in an embodiment of the present invention; Figure 2 is a schematic diagram of the system module structure provided in an embodiment of the present invention; Figure 3 is a schematic diagram of spatial accessibility information provided in an embodiment of the present invention; Figure 4 is a schematic diagram of the physical constraint parameters of the main body provided in an embodiment of the present invention; Figure 5 is a schematic diagram of the entrance pose, movable area and runtime control provided in an embodiment of the present invention; Figure 6 is a schematic diagram of the existing space accessibility information verification, fusion or reconstruction process provided in the embodiments of the present invention. Detailed Implementation

[0025] The present invention will be described below with reference to embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0026] (a) Terminology Explanation A 3D spatial representation refers to a computable representation used to describe the geometry, structure, field, point cloud, mesh, semantics, navigation, collision, or other spatial information of 3D content. A 3D spatial representation may include one or more of the following: mesh model, point cloud, voxel, symbolic spatial description, spatial field, neural 3D representation, extended reality scene understanding results, artificial intelligence-generated spatial results, or spatial structure information built into a 3D spatial format.

[0027] An object is an entity that needs to enter three-dimensional content and be presented, observed, moved, docked, or perform tasks within it. An object can be a human user, a digital entity corresponding to a human user, a digital human, a robot, an autonomous vehicle, a drone, a virtual camera, or other embodied objects.

[0028] The physical constraint parameters of a subject refer to the technical parameters characterizing its geometric dimensions, viewpoint, envelope, movement, turning, traversal, or docking capabilities. These parameters may include one or more of the following: viewpoint height, subject envelope size, movement radius, turning radius, traversable height, minimum accessible width, dockable area, subject movement speed range, or actuator boundaries. These parameters can be input by the user or determined by system default configurations, built-in constants, configuration files, subject type mappings, device capabilities, sensor information, historical behavior, or runtime detection results. Obtaining these physical constraint parameters may involve reading, calling, loading, selecting, or determining these parameters.

[0029] Spatial accessibility information refers to three-dimensional spatial data used to characterize whether an entity can enter, stand, stop, move, or be restricted by boundaries. Spatial accessibility information may include one or more of the following: support area, clear area, passable area, boundary constraints, candidate entrance, navigation area, collision boundary, spatial semantics, occupancy field, distance field, or navigation grid.

[0030] Support regions and clearance regions can be represented by explicit region data, or implicitly by navigation grids, occupancy fields, distance fields, collision detection rules, ray detection rules, capsule detection rules, voxel occupancy states, or other spatial data structures. Any data or rules that can be used to determine the position where the subject can stand, dock, or move, and to determine the collision-free space of the subject's envelope in the height or volume direction, can be used as a representation of the support region or clearance region.

[0031] The entrance pose refers to the initial spatial state of a subject when it enters 3D content. It can include one or more of the following: entrance position, entrance orientation, camera initial pose, subject initial pose, entry point, spawn point, or viewpoint.

[0032] The actionable area refers to the area in the runtime environment in which an entity is allowed to move, dock, or perform actions.

[0033] An accessible area refers to the area that meets the entry conditions of the subject within the accessibility result of the thematic space. The accessible area can be determined by at least the support area and the clearance area, and can be further determined by combining the passable area or boundary constraints, and is used to determine the entrance location.

[0034] Physical boundary constraints refer to data or rules used to restrict an entity from traversing obstacles, collision boundaries, ground boundaries, or spatial boundaries. Physical boundary constraints can be explicit boundary data or implicitly represented by actionable area boundaries, navigation mesh boundaries, colliders, occupancy fields, distance fields, ray detection rules, or capsule detection rules.

[0035] The runtime environment refers to the execution environment that controls the camera pose or the subject's movement boundaries based on the entrance pose, the movable area, and physical boundary constraints. The runtime environment can be a mobile 3D rendering environment, a Web 3D environment, a 3D runtime engine, an extended reality device, a digital twin platform, a robot simulation system, or a robot control system.

[0036] Identity, permissions, task objectives, or device differences can serve as optional contexts for selecting or modifying subject physical constraint parameters, entrance orientation, actionable area, or boundary constraints, but do not replace spatial calculations and subject physical constraint verification.

[0037] (II) Overall Process Implementation Example As shown in Figure 1, in one embodiment, the method for determining the accessibility of a three-dimensional space and generating an immersive entrance based on the physical constraints of the subject includes steps S101 to S106.

[0038] S101, Obtain the three-dimensional spatial representation.

[0039] The three-dimensional space representation is used to describe the three-dimensional content to be entered. The three-dimensional content can be a room, exhibition hall, street, museum space, digital city area, AI-generated space, real-world scanned space, or other three-dimensional content. The system uses spatial accessibility information to determine whether the three-dimensional content possesses spatial conditions suitable for a subject to enter or move within. The three-dimensional space representation can include one or more of the following: geometric data, point cloud data, mesh data, voxel data, spatial field data, symbolic spatial structure, navigation information, collision information, or spatial semantic information.

[0040] S102, Obtain the physical constraint parameters of the main body.

[0041] The physical constraint parameters of the subject are used to characterize the geometric scale or movement constraints of the subject entering the 3D content. As shown in Figure 4, the physical constraint parameters of the subject may include parameters such as viewpoint height, subject envelope size, movement radius, turning radius, and traversable height. Obtaining the physical constraint parameters of the subject may include reading, loading, or determining the physical constraint parameters from user input, system default configuration, built-in constants, configuration files, subject type mapping tables, device capability information, sensor information, historical records, or runtime detection results. For ordinary adult users, the physical constraint parameters of the subject may include the adult default viewpoint height, default subject envelope size, and default movement radius; for child users, they may include a lower viewpoint height and a more conservative movement boundary; for wheelchair users, they may include the wheelchair envelope size, a larger turning radius, and a smaller traversable height; for robot subjects, they may include the robot body envelope size, turning radius, traversable height, and sensor field of view.

[0042] S103, perform spatial calculations on the three-dimensional spatial representation to generate spatial accessibility information; or, if the three-dimensional spatial representation or its associated data already contains spatial accessibility information, read or receive the spatial accessibility information.

[0043] In the generation branch, the system can perform geometric analysis, normal vector analysis, bounding box analysis, occupancy field sampling, distance field calculation, collision detection, connected component analysis, or navigation region calculation on the 3D spatial representation, obtaining at least the support region and the clearance region, and can further obtain the traversable region or boundary constraints. As shown in Figure 3, the support region is used to determine the position where the subject can stand or stop, the clearance region is used to determine the collision-free space of the subject's envelope in the height or volume direction, the traversable region is used to determine the range of movement of the subject, and the boundary constraints are used to restrict the subject from crossing obstacles or spatial boundaries.

[0044] In some embodiments, the support region and clearance region are not required to be output as separately named data layers. The system can implicitly represent the support region and clearance region through the navigation mesh, occupancy field, distance field, collider, voxel occupancy state, or runtime collision rules. For example, when the navigation mesh generation process has excluded unstandable positions and positions without clearance based on the subject's height, subject radius, slope, elevation difference, or collision detection results, the navigation mesh or its generation result can serve as an implicit representation of the support region and clearance region.

[0045] In the reading or receiving branch, if the 3D spatial representation, 3D format, external engine, extended reality operating system, artificial intelligence generation system, or manual annotation results already contain candidate entrances, navigation areas, collision boundaries, spatial semantics, walkable areas, spawn points, or other spatial accessibility information, the system can read or receive this information and proceed to the subsequent verification, fusion, or reconstruction process.

[0046] The aforementioned generation and reading / receiving branches do not use the source of spatial accessibility information as the basis for distinguishing the technical contributions of this invention. Regardless of whether the spatial accessibility information is calculated by the system or provided by a 3D format, external engine, extended reality operating system, artificial intelligence generation system, or manual annotation results, the system combines the subject's physical constraint parameters to perform subjectification processing on the spatial accessibility information and generate the entrance pose, movable area, and physical boundary constraints.

[0047] S104. Based on the physical constraint parameters of the subject, the spatial accessibility information is verified, fused, or reconstructed to obtain a subjectified spatial accessibility result that matches the subject.

[0048] In some embodiments, the system verifies whether there is sufficient clearance at the candidate entrance, whether the candidate entrance meets the body envelope size, whether the navigation area meets the body's movement radius or turning radius, whether the collision boundary can restrict the body from crossing obstacles, and whether the spatial semantics are consistent with the spatial calculation results.

[0049] When spatial accessibility information comes from multiple sources, the system can fuse information from multiple sources. When existing spatial accessibility information is incomplete, inconsistent, or does not meet the main physical constraint parameters, the system can reconstruct at least a portion of the spatial accessibility information based on the three-dimensional spatial representation.

[0050] In some embodiments, when existing spatial accessibility information already meets the integrity and consistency requirements, the system can perform the verification step separately and output the subject-based spatial accessibility result; when spatial accessibility information from multiple sources is available but there are differences in accuracy, confidence, or coverage, the system can perform the fusion step separately and output the subject-based spatial accessibility result; when existing spatial accessibility information is missing or insufficient to meet the subject physical constraint parameters, the system can perform the reconstruction step separately and output the subject-based spatial accessibility result.

[0051] S105 generates the entrance pose, movable area, and physical boundary constraints based on the subjectified space accessibility results.

[0052] As shown in Figure 5, the entrance pose can include the entrance location and the entrance orientation. The entrance location can be set within an accessible area that matches the physical constraint parameters of the main body. The entrance orientation is primarily determined based on the open spatial direction. When the open spatial direction does not meet the preset conditions, it can be determined based on the main spatial axis or candidate entrances. When there are task objectives or boundary constraints, the entrance orientation can be corrected based on the task objectives or boundary constraints.

[0053] The movable area is used to limit the range of movement of the subject. Physical boundary constraints are used to restrict the subject from crossing obstacle boundaries, collision boundaries, or ground boundaries. Physical boundary constraints can be implemented by explicit colliders, navigation area boundaries, occupancy fields, distance fields, capsule detection rules, ray detection rules, or other movement constraint strategies.

[0054] S106 outputs the entrance pose, movable area, and physical boundary constraints so that the runtime environment can control the camera pose or the subject's movement boundary according to the above parameters, and import the subject into the 3D content.

[0055] The output may be sent to a runtime module within the same device, or it may be written to a storage medium, sent to the cloud or terminal via a network, returned via an application programming interface, or provided to a 3D runtime engine, extended reality device, robot simulation system, or robot control system.

[0056] When the subject is a human user, the runtime environment can set the initial position and orientation of the first-person camera based on the entry pose, and restrict the user's movement boundaries based on the movable area and physical boundary constraints. When the subject is a robot or digital subject, the runtime environment can set the subject's initial pose based on the entry pose, and restrict the subject's range of motion based on the movable area and physical boundary constraints.

[0057] (III) Example of generating spatial accessibility information As shown in Figure 3, in one embodiment, the system generates spatial accessibility information based on a three-dimensional spatial representation.

[0058] First, the system preprocesses the 3D spatial representation. This preprocessing may include reading the model bounding box, unifying coordinate axis orientation, estimating ground height, estimating unit scale, removing outlier geometry, or generating sampling points.

[0059] Secondly, the system detects support regions. Support regions can be determined through normal vector filtering, slope thresholding, horizontal plane detection, point cloud plane fitting, voxel bottom surface detection, or semantic labeling. Support regions are used to determine candidate locations where the subject can stand, stop, or move.

[0060] Next, the system calculates the clearance area. The clearance area can be obtained by collision detection of the subject's envelope along the height or volume direction. For example, when the subject is a human user, a capsule or bounding box can be used to simulate the human body's envelope; when the subject is a robot, the robot's body envelope can be used. If the subject's envelope collides with an obstacle in the three-dimensional spatial representation at a certain location, then that location does not meet the clearance condition.

[0061] The system then calculates the passable area. The passable area can be obtained by sampling the support area and performing main envelope collision detection, connected component analysis, movement radius detection, turning radius detection, or traversable height detection at the sampling points.

[0062] Finally, the system generates boundary constraints. Boundary constraints can consist of obstacle boundaries, collision boundaries, ground boundaries, wall boundaries, spatial boundaries, navigation area boundaries, or occupied area boundaries. Boundary constraints are used to restrict an entity from traversing obstacles or leaving permitted access areas in the runtime environment.

[0063] (iv) Example of verifying accessibility information of existing space As shown in Figure 6, in one embodiment, the 3D spatial representation or its associated data already includes entry-level information or boundary-level information. Entry-level information may include candidate entry points, preset entry points, spawn points, navigation areas, or walkable areas. Boundary-level information may include collision boundaries, spatial semantics, wall boundaries, obstacle areas, occupied areas, or navigation boundaries.

[0064] The system does not directly use the existing information mentioned above, but instead performs verification based on the main physical constraint parameters.

[0065] The verification may include verifying whether the entry information meets the subject entry conditions and whether the boundary information meets the subject movement boundary conditions. The subject entry conditions may include at least one of the following: the entry location is within the support area, the entry location has clearance, the entry location matches the subject's envelope size, and the entry orientation faces the available space. The subject movement boundary conditions may include boundary information that restricts the subject from crossing at least one of the following: obstacle boundary, collision boundary, ground boundary, or spatial boundary.

[0066] When the existing entry-level information, navigation area information, and boundary information are complete and consistent, verification can be used as an independent processing step. In this case, the system verifies whether the above information meets the subject's physical constraint parameters to obtain a subject-specific spatial accessibility result that matches the subject, without having to regenerate all spatial accessibility information.

[0067] For candidate entrances, the system can verify whether the entrance location is within the support area, whether there is sufficient clearance at the entrance location, whether the entrance location matches the size of the main body envelope, and whether the entrance orientation faces the available space rather than a wall or obstacle.

[0068] For the navigation area, the system can verify whether the width of the navigation area meets the main body's movement radius or the main body's envelope size, verify whether the turning area meets the main body's turning radius, and verify whether the steps or slopes meet the height that the main body can cross.

[0069] Regarding collision boundaries, the system can verify whether the collision boundary covers the boundaries of walls, display cases, furniture, building facades, or the ground, and whether it can restrict the subject from passing through the obstacle.

[0070] Regarding spatial semantics, the system can verify whether the spatial semantics are consistent with the geometric calculation results. For example, if a region is marked as ground, but its normal vector, slope, or height does not meet the support region conditions, the system can reduce the confidence of the semantic result or trigger reconstruction.

[0071] (v) Integration or Reconstruction of Implementation Examples As shown in Figure 6, in one embodiment, spatial accessibility information comes from multiple sources. For example, a 3D model format can have built-in candidate entrances, an external engine can generate navigation areas, an extended reality system can provide wall and ground semantics, and the system can generate support areas and clearance areas through geometric calculations.

[0072] The system can fuse results from various sources based on the main physical constraint parameters. Fusion may include weighted scoring of candidate entry points, finding the intersection or union of support and navigation regions, performing consistency checks on collision and semantic boundaries, and prioritizing boundaries from different sources.

[0073] When multiple sources provide spatial accessibility information, but these sources differ in spatial coverage, accuracy, confidence level, or semantic granularity, fusion can be used as an independent processing step. The system can select, weight, or combine the results from each source based on the subject's physical constraint parameters to obtain a subject-specific spatial accessibility result.

[0074] When existing spatial accessibility information is incomplete, inconsistent, or fails verification, the system can reconstruct at least a portion of the spatial accessibility information based on the 3D spatial representation. For example, when there are no candidate entrances in the model, the system can select entrance candidates within the maximum passable area; when collision boundaries are missing, the system can generate physical boundary constraints based on geometric boundaries, occupancy fields, or distance fields; when the navigation area does not meet the turning radius of the main body, the system can reduce the traversable area or regenerate the traversable area.

[0075] When existing space accessibility information is missing, unreadable, or clearly does not meet the main physical constraint parameters, reconstruction can be used as an independent processing step. The system can recalculate at least a portion of the support region, clearance region, passable region, or boundary constraints based on the three-dimensional spatial representation to obtain the subjectified space accessibility results.

[0076] (vi) Example of entrance pose generation As shown in Figure 5, in one embodiment, the system generates the entrance pose based on the accessibility results of the subjectified space.

[0077] The entrance location can be selected from the accessible area that meets the physical constraints of the main body. The system can prioritize locations that are within the largest connected and passable area, maintain a safe distance from obstacles, have sufficient clearance, and will not cause collision with the initial envelope of the main body.

[0078] The entrance orientation can be primarily determined based on the open spatial direction. This open spatial direction can be obtained by emitting detection lines from candidate entrance locations in multiple directions, detecting collision-free distances, calculating obstacle density, calculating the visible area, or analyzing the extension direction of the passable area. When the open spatial direction does not meet preset conditions, the entrance orientation can be determined based on the spatial principal axis or candidate entrances. These preset conditions may include at least one of the following: a maximum collision-free distance greater than a preset distance threshold, a visible area greater than a preset area threshold, obstacle density lower than a preset density threshold, or a direction confidence level higher than a preset confidence threshold. The spatial principal axis can be determined using the bounding box major axis, principal component analysis, the connected region major axis, or the navigation region direction. When there are task objectives or boundary constraints, the system can correct the entrance orientation based on these objectives or constraints.

[0079] In some embodiments, the entry pose can be generated using one or more of the following modes: automatic generation mode, candidate selection mode, recommendation confirmation mode, or interactive assistance mode.

[0080] In automatic generation mode, the system calculates the entrance position and orientation based on the accessibility results of the subjectified space and directly outputs the entrance pose. This entrance pose can be used by the runtime environment to set the initial pose of the camera or the starting pose of the subject.

[0081] In the candidate selection mode, the system generates multiple candidate entrance poses based on the accessibility results of the subjectified space, and can sort these candidate entrance poses according to safety distance, clearance margin, area of ​​connected regions, open space direction, boundary constraints, or mission objectives. Users, runtime environments, or other system modules can select one entrance pose from these multiple candidate entrance poses, and the system can fine-tune the position, orientation, or boundary constraints based on the selection result.

[0082] In the recommendation confirmation mode, the system generates a recommended entrance pose based on the accessibility results of the subjectified space and provides the recommended entrance pose to the user or other system modules for confirmation. The user or other system modules can accept the recommended entrance pose or adjust the entrance location, entrance orientation, movable area, or physical boundary constraints; the system can continue to perform subject entry condition or subject movement boundary condition verification on the adjusted results.

[0083] In interactive assistance mode, users can specify, move, or adjust the entrance location or orientation within the 3D content. The system provides verification feedback based on the accessibility results of the thematic space. This verification feedback can include whether the entrance location is within an accessible area, whether there is sufficient clearance, whether it meets the subject's envelope dimensions, whether it collides with obstacle boundaries, whether it is within a movable area, or whether it meets physical boundary constraints. The system can also correct, snap, or project the user-specified location to an accessible area that meets the subject's physical constraint parameters.

[0084] All of the above generation modes determine the entrance pose, movable area, or physical boundary constraints based on the subject-based spatial accessibility results. User confirmation, user selection, or user adjustment do not change the technical correlation between the entrance pose, the subject-based physical constraint parameters, and the spatial accessibility information.

[0085] (vii) Examples of movable areas and physical boundary constraints As shown in Figure 5, in one embodiment, the system generates an actionable region and physical boundary constraints based on the accessibility results of the subjectified space.

[0086] The movable area can include areas where the subject can move or dock. The movable area can be calculated from the support area, clearance area, passable area, navigation area, permission area, or task area.

[0087] Physical boundary constraints can include explicit or implicit boundary data. Explicit boundary data can include colliders, wall boundaries, obstacle boundaries, ground boundaries, movable area boundaries, or navigation mesh boundaries. Implicit boundary data can include occupancy fields, distance fields, collision detection rules, ray detection rules, capsule detection rules, or runtime movement constraint strategies.

[0088] The runtime environment can control the movement boundaries of a subject based on the movable area and physical boundary constraints. For example, when a subject attempts to move towards a wall, the runtime environment can prevent the subject from passing through the wall through collision detection or movement boundary detection; when a subject attempts to leave the ground boundary, the runtime environment can prevent movement or trigger a degraded mode; when a subject approaches a display case, furniture, or other obstacle, the runtime environment can restrict the subject from continuing to move forward.

[0089] (viii) Calculation Implementation Example In one embodiment, the system may generate the subject-based space accessibility result using the following calculation method. The following calculation method is used to illustrate feasible technical paths and does not constitute a limitation on specific algorithms, thresholds, or data structures.

[0090] The system can represent the three-dimensional space as M and denote the physical constraint parameters of the main body as P, where P includes at least one of the following: viewpoint height h, main body envelope radius r, main body height H, turning radius R, or traversable height c. The system can generate a set of sampling points X in M ​​and calculate support conditions, clearance conditions, traversability conditions, and boundary conditions for each sampling point x.

[0091] In one example, the support condition Support(x) can be determined based on the surface normal vector near the sampling point, the slope, or the plane fitting result. For example, when the angle between the surface normal vector near the sampling point and the vertical direction is less than a preset slope threshold, and the height change of the sampling point is less than a preset height difference threshold, the sampling point is determined to meet the support condition.

[0092] In one example, the clearance condition Clearance(x, P) can be determined based on the collision detection results between the main body envelope and the 3D spatial representation M. For example, a capsule, cylinder, or bounding box corresponding to the main body's physical constraint parameters P is constructed with the sampling point x as the bottom position; the sampling point is determined to satisfy the clearance condition if the envelope does not intersect with the obstacle geometry, occupied voxel, or collision boundary in M ​​within the space corresponding to height H or viewpoint height h.

[0093] In one example, the traversable condition `Traversable(x, P)` can be determined based on support conditions, clearance conditions, connectivity, and body movement constraints. For instance, if a sampling point `x` satisfies `Support(x)` and `Clearance(x, P)`, and the traversable width between `x` and at least one adjacent sampling point is not less than the body envelope size or movement radius `r`, and the turning area satisfies the turning radius `R`, then `x` is marked as a traversable sampling point. Connectivity analysis of the traversable sampling points yields one or more traversable regions.

[0094] In one example, the accessibility result of the subjectified space can be represented as: E(P, M) = { x | Support(x) = true, Clearance(x, P) = true, Traversable(x, P) = true} (1) Where E(P, M) represents the set of accessible or passable regions that match the physical constraint parameter P of the main body. The above expression is only used to illustrate the calculation relationship of the physical constraint parameter of the main body in spatial accessibility determination. In actual implementation, meshes, point clouds, voxels, occupancy fields, distance fields, navigation meshes or other spatial data structures can be used.

[0095] In one example, the system can generate a candidate entry point set C based on E(P, M) and score the candidate entry point e. The score can be a combination of one or more of the following: safety distance, clearance margin, area of ​​connected regions, open space direction, distance to obstacles, semantic consistency with the candidate entry point, or direction of the task objective. For example, the candidate entry point score Score(e) can be represented as: Score(e) = w1 × d(e) + w2 × a(e) + w3 × o(e) + w4 × s(e) (2) Where d(e) represents the distance or safety margin from the candidate entrance to the obstacle boundary, a(e) represents the area of ​​the connected region or the movable range where the candidate entrance is located, o(e) represents the openness corresponding to the entrance orientation, s(e) represents the consistency between the candidate entrance and existing spatial semantics or candidate entrance information, and w1 to w4 are weight parameters. The system can select candidate entrances that meet the preset scoring conditions to generate entrance poses.

[0096] In one example, physical boundary constraints can be generated based on the movable region A and the obstacle region O. When the runtime environment receives a subject movement request, it can detect whether the moved subject envelope B(P, q) is still within the movable region A and whether it intersects with the obstacle region O; where q represents the runtime pose of the subject or camera. When B(P, q) exceeds A or intersects with O, the runtime environment can reject the movement request, correct the movement vector, or project the subject pose back into the allowed region.

[0097] (ix) Human User Implementation Examples In one embodiment, a user uploads or generates a 3D spatial model of a room. The system obtains this 3D spatial representation and acquires the user's corresponding physical constraint parameters, such as viewpoint height, subject envelope size, and movement radius.

[0098] The system performs spatial calculations on this 3D spatial representation, detects the ground or other supporting areas, calculates the clearance area within the user's viewpoint height range, and generates passable areas and boundary constraints. If the model or associated data already contains candidate entrances or collision boundaries, the system verifies them based on the main physical constraint parameters.

[0099] The system generates entrance pose, movable area, and physical boundary constraints based on the subject-based spatial accessibility results. The runtime environment sets the position and orientation of the first-person camera according to the entrance pose and restricts user movement according to the movable area and physical boundary constraints, enabling users to enter the three-dimensional space of the room, rather than just observing the model from the outside.

[0100] When the user is a child or a user with a low field of vision, the system can adopt a lower viewpoint height and a more conservative movement boundary; when the user is a wheelchair user, the system can adopt a larger body envelope size and turning radius, and adjust the movable area.

[0101] In another embodiment, the 3D content can be 3D spatialized content constructed based on user images, videos, text, event logs, location logs, or other historical content. The 3D spatialized content can be generated from user-uploaded materials, scanned data, artificial intelligence-generated results, or 3D modeling results. The system can generate an entry pose, movable area, and physical boundary constraints for the user to enter the 3D spatialized content based on the 3D spatial representation corresponding to the 3D spatialized content and the user's physical constraint parameters. This allows the user to observe, move, or interact within the 3D spatialized content in a manner matching their physical constraint parameters.

[0102] (x) Digital Museum Implementation Examples In one embodiment, the 3D content is a digital museum exhibition hall. The exhibition hall model may include display cases, walls, entrances, exhibition lines, walkable areas, and restricted areas.

[0103] The system acquires the 3D spatial representation of the exhibition hall and the main physical constraint parameters. For ordinary visitors, the system generates the pose of the public entrance, the public movable area, and the physical boundary constraints of the display cases; for tour guides, the system can generate the entrance orientation facing the tour route; for administrators, the system can modify the movable area according to permissions or task objectives, enabling them to enter the maintenance area.

[0104] In the above process, permissions and identity do not determine the entrance independently, but can be used to modify the movable area or the entrance orientation. The system still verifies the entrance location, clearance, traversable area, and boundary constraints based on the main physical constraint parameters.

[0105] (xi) Examples of Digital Cities In one embodiment, the 3D content is a digital street block or digital city space. The subjects can be pedestrians, vehicles, robots, or drones. Different subjects have different physical constraint parameters.

[0106] For pedestrians, the system can generate a movable area based on the human viewpoint height, the size of the subject's envelope, and the movement radius; for vehicles, the system can generate a movable area based on the vehicle size, turning radius, and road boundaries; and for robots, the system can generate the initial pose and movement boundaries based on the robot's envelope size, traversable height, and sensor field of view.

[0107] (xii) Robot Body Embodiment In one embodiment, the subject is a robot, and the three-dimensional space is represented as the robot's task space or digital twin space. The subject's physical constraint parameters include the robot's body envelope dimensions, turning radius, traversable height, dockable area, and sensor field of view.

[0108] The system performs spatial calculations on the 3D spatial representation to obtain the support region, clearance region, traversable region, and boundary constraints. Based on the robot's physical constraint parameters, the system verifies whether candidate entrances or navigation regions allow the robot to enter, and verifies whether collision boundaries can restrict the robot's obstacle crossing. If the existing navigation region does not meet the robot's turning radius or clearance conditions, the system can reconstruct the traversable region.

[0109] The system generates the robot's initial pose, movable area, and physical boundary constraints for entering 3D space or performing tasks. The runtime environment can be a robot simulation system or a robot control system, used to control the robot's initial pose and movement boundaries.

[0110] (xiii) Extending Real-World Scenarios to Understand Input Examples In one embodiment, the representation of the three-dimensional space or its associated data is provided by an extended reality system. The extended reality system can provide ground, walls, spatial anchor points, room boundaries, collision meshes, or scene semantics.

[0111] This invention can read or receive the aforementioned spatial accessibility information and verify it based on the physical constraint parameters of the subject. For example, the room boundaries provided by an extended reality system may be suitable for an average adult user, but not necessarily for a robot or a wheelchair user. This invention can verify whether the boundary is suitable for the current subject based on the subject's envelope size, movement radius, or turning radius.

[0112] When the information provided by the extended reality system is incomplete or inconsistent, the present invention can integrate information from multiple sources or reconstruct at least a portion of spatial accessibility information based on a three-dimensional spatial representation.

[0113] (XIV) Example of Artificial Intelligence Generated Spatial Input In one embodiment, the 3D spatial representation is generated by an artificial intelligence (AI) generation system. This AI generation system can output a mesh model, spatial semantics, materials, colliders, navigation regions, candidate entrances, or other spatial structure information.

[0114] This invention can use the information output by the artificial intelligence generation system as one of the sources of spatial accessibility information and verify it based on the physical constraint parameters of the subject. If the candidate entrance generated by the artificial intelligence is located inside an obstacle, lacks clearance, or does not meet the subject envelope size, the system can reject the entrance, integrate information from other sources, or reconstruct a new entrance pose, movable area, and physical boundary constraints based on the three-dimensional spatial representation.

[0115] Therefore, even if future 3D generation systems are able to output semantics, colliders, navigation regions, or spawn points, this invention can still perform subjectification verification and runtime control parameter generation on this information.

[0116] (xv) Downgraded and modified implementation examples In one embodiment, when the confidence level of the accessibility result of the subjectified space is lower than a preset condition, the system can output an external preview mode, a semi-immersive preview mode, or a manually confirmed entry mode. This processing is an optional anomaly handling mechanism and is not a necessary step in generating the entry pose, movable area, and physical boundary constraints. For example, when the 3D spatial representation geometry is severely incomplete, the support area cannot be reliably determined, the clearance area cannot be verified, or the boundary constraints are incomplete, the system may not directly import the subject into the 3D space, but instead first enter an external preview or request user confirmation.

[0117] In another embodiment, the user can manually modify the entrance pose, movable area, or boundary constraints. The system can record this modification and use the result when generating subsequent entrance poses, movable areas, or physical boundary constraints. This modification can be used to reduce the risk of subsequent invalid entrances or boundary crossings.

[0118] (xvi) Relationship with spatial normalization technology In some embodiments, the present invention can be used in conjunction with cross-platform 3D spatial data interoperability or spatial normalization techniques. Spatial normalization techniques can convert 3D spatial data from different sources, coordinate systems, scales, or platforms into a unified or computable 3D spatial representation.

[0119] This invention does not repeatedly solve the coordinate normalization problem of spatial data from different platforms. Instead, based on a unified or computable three-dimensional spatial representation, it further combines the physical constraint parameters of the subject to verify, fuse, or reconstruct the spatial accessibility information and generate entrance pose, movable area, and physical boundary constraints to control the camera pose or subject movement boundary during runtime. System Implementation Examples

[0120] As shown in Figure 2, in one embodiment, the three-dimensional space accessibility judgment and immersion entrance generation system based on subject physical constraints includes a spatial representation acquisition module, a subject constraint acquisition module, a spatial calculation module, a subjectification processing module, an entrance generation module, and a runtime control module.

[0121] The spatial representation acquisition module is used to acquire a 3D spatial representation. This 3D spatial representation can come from local files, cloud interfaces, scanning devices, 3D editors, AI-generated systems, extended reality systems, or 3D format parsers.

[0122] The subject constraint acquisition module is used to obtain the subject's physical constraint parameters. These parameters can be determined by user input, system default configuration, built-in constants, configuration files, device configuration, subject type mapping, sensor measurements, or historical records.

[0123] The spatial computation module is used to perform spatial calculations on the 3D spatial representation and generate spatial accessibility information. The spatial computation module can also read or receive spatial accessibility information when the 3D spatial representation or its associated data already contains such information.

[0124] The subjectification processing module includes a verification unit and a fusion and reconstruction unit. The verification unit verifies the spatial accessibility information based on the subject's physical constraint parameters after the spatial computing module reads or receives the information. The fusion and reconstruction unit fuses spatial accessibility information from multiple sources, or reconstructs at least a portion of the spatial accessibility information based on a 3D spatial representation, when verification fails, or when the spatial accessibility information is incomplete or inconsistent, to obtain a subjectified spatial accessibility result that matches the subject.

[0125] The entrance generation module is used to generate entrance pose, movable area and physical boundary constraints based on the subjectified space accessibility results.

[0126] The runtime control module is used to control the camera pose or the subject's movement boundary based on the entrance pose, the movable area, and physical boundary constraints.

[0127] The modules described above can be implemented using software, hardware, or a combination of both. These modules can be deployed on the same device or distributed across mobile devices, web browsers, the cloud, edge computing, extended reality devices, 3D runtime engines, robot bodies, or robot control systems. The modules are logically divided; physical independence is not required. Storage Media Examples

[0128] In one embodiment, the present invention also provides a computer-readable storage medium. This computer-readable storage medium stores a computer program or instructions that, when executed by a processor, implement the method for determining 3D spatial accessibility and generating an immersive entrance based on subject physical constraints as described in any of the above embodiments.

[0129] Computer-readable storage media may include mobile device memory, server memory, edge device memory, extended reality device memory, robot controller memory, solid-state drives, flash memory, read-only memory, random access memory, or other media capable of storing computer programs or instructions. Optional Deformation

[0130] The above embodiments are merely illustrative examples of the present invention. Those skilled in the art can substitute or combine the sources of the three-dimensional spatial representation, the methods for obtaining the main physical constraint parameters, the methods for generating spatial accessibility information, the methods for verification fusion or reconstruction, the methods for generating entrance poses, and the methods for implementing movable areas and physical boundary constraints without departing from the spirit of the present invention.

[0131] For example, 3D spatial representation can be achieved by meshes, point clouds, voxels, SDF, neural 3D representations, or extended reality scene understanding results; physical boundary constraints can be achieved by colliders, navigation meshes, occupancy fields, distance fields, or runtime collision detection rules; entrance pose can be used for first-person perspective, third-person perspective, free camera, digital subject, or robot initial pose.

[0132] As long as the technical solution is still based on three-dimensional spatial representation and physical constraint parameters of the subject, and verifies, fuses or reconstructs spatial accessibility information, and generates entrance pose, movable area and physical boundary constraints to control the camera pose or subject movement boundary during operation, it should fall within the protection scope of this invention.

Claims

1. A method for determining the accessibility of a three-dimensional space and generating an immersive entrance based on the physical constraints of the subject, characterized in that, include: Obtain a three-dimensional spatial representation, which is used to describe the three-dimensional content to be entered; Obtain the physical constraint parameters of the subject, which are used to characterize the geometric scale or action constraints of the subject to enter the three-dimensional content; Spatial calculations are performed on the three-dimensional spatial representation to generate spatial accessibility information; If the three-dimensional spatial representation or its associated data already contains spatial accessibility information, then the spatial accessibility information is read or received; wherein, the spatial accessibility information is used to characterize the spatial conditions in the three-dimensional content that allow the subject to enter or move, and the spatial accessibility information includes at least one of a support area and a clear area, as well as a passable area or a boundary constraint; Based on the physical constraint parameters of the subject, the spatial accessibility information is verified, fused, or reconstructed to obtain a subjectified spatial accessibility result that matches the subject. Based on the subjectified space accessibility results, the entrance pose, movable area, and physical boundary constraints are generated. The entry pose, movable area, and physical boundary constraints are output so that the runtime environment controls the camera pose or the subject movement boundary according to the entry pose, movable area, and physical boundary constraints, and imports the subject into the 3D content.

2. The method according to claim 1, characterized in that, The subject is a human user or a digital subject corresponding to a human user; the entry pose is used to enable the human user to enter the three-dimensional content from a first-person perspective, and to match the first-person perspective with the physical constraint parameters of the subject.

3. The method according to claim 1, characterized in that, The physical constraint parameters of the subject include at least one of the following: viewpoint height, subject envelope size, movement radius, turning radius, or traversable height.

4. The method according to claim 1, characterized in that, The spatial accessibility information includes a support area, a clearance area, a passable area, and boundary constraints; wherein, the support area is used to determine the position where the subject can stand or stop, the clearance area is used to determine the collision-free space of the subject's envelope in the height or volume direction, the passable area is used to determine the range of movement of the subject, and the boundary constraints are used to restrict the subject from crossing obstacles or spatial boundaries.

5. The method according to claim 1, characterized in that, If the three-dimensional spatial representation or its associated data already contains entry-type information or boundary-type information, verification is performed based on the physical constraint parameters of the subject; wherein, the entry-type information includes candidate entry points or navigation areas, and the boundary-type information includes collision boundaries or spatial semantics; the verification includes verifying whether the entry-type information satisfies the subject entry conditions, and verifying whether the boundary-type information satisfies the subject movement boundary conditions.

6. The method according to claim 1, characterized in that, When existing spatial accessibility information is incomplete, inconsistent, or does not meet the main physical constraint parameters, spatial accessibility information from multiple sources is fused, or at least a portion of the spatial accessibility information is reconstructed based on the three-dimensional spatial representation.

7. The method according to claim 1, characterized in that, The entrance pose includes the entrance position and the entrance orientation. The entrance position is set in an accessible area that matches the physical constraint parameters of the main body. The entrance orientation is determined primarily based on the open space direction. When the open space direction does not meet the preset conditions, it is determined based on the main axis of space or a candidate entrance. When there are task objectives or boundary constraints, the entrance orientation is corrected based on the task objectives or boundary constraints.

8. The method according to claim 1, characterized in that, The movable area and the physical boundary constraints are used to control the movement boundaries of the subject in the three-dimensional content, so that the subject is restricted to the area that can be entered in the runtime environment, and avoids crossing at least one of the obstacle boundary, collision boundary or ground boundary.

9. A three-dimensional space accessibility judgment and immersion entrance generation system based on subject physical constraints, characterized in that, include: The spatial representation acquisition module is used to acquire a three-dimensional spatial representation; The main constraint acquisition module is used to acquire the physical constraint parameters of the main body; The spatial computing module is used to perform spatial calculations on the three-dimensional spatial representation to generate spatial accessibility information, or to read or receive the spatial accessibility information when the three-dimensional spatial representation or its associated data already contains spatial accessibility information. The subject-based processing module includes a verification unit and a fusion and reconstruction unit; the verification unit is used to verify the space accessibility information according to the subject physical constraint parameters after the space computing module reads or receives the space accessibility information. The fusion and reconstruction unit is used to fuse spatial accessibility information from multiple sources or reconstruct at least a portion of spatial accessibility information based on the three-dimensional spatial representation when verification fails or spatial accessibility information is incomplete or inconsistent, so as to obtain a subjectified spatial accessibility result that matches the subject. The entrance generation module is used to generate the entrance pose, movable area and physical boundary constraints based on the subjectified space accessibility results. The runtime control module is used to control the camera pose or the subject movement boundary based on the entrance pose, the movable area, and the physical boundary constraints.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program or instructions that, when executed by a processor, implement the method according to any one of claims 1 to 8.

Images