Bone animation interactive display method
By generating standardized joint trees, integrating multi-source data, and building a multimodal interactive closed loop, the problems of continuity and accuracy in fast movements and detail adjustments of skeletal animation are solved, achieving efficient production and smooth display of skeletal animation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-25
- Publication Date
- 2026-03-31
AI Technical Summary
In existing technologies, skeletal animation has good continuity during fast movements, but it is difficult to adjust details. When users operate it manually, it is easy to cause stuttering and stiffness, making it difficult to meet the animation production and interaction needs in complex scenes.
Standardized hierarchical joint trees are generated using multi-view images or line drawings. Combined with dynamic isolation of conflict areas and weight self-optimization algorithms, physical motion capture, text commands and user operation data are integrated. The joint trees are driven by a multi-source data spatiotemporal collaborative fusion engine to build a multimodal interactive closed loop. The cloud collaborative platform supports simultaneous editing by multiple people, forming a complete closed loop.
It significantly improves the smoothness of motion and the accuracy of interaction in skeletal animation, solving the problem of balancing speed and accuracy in complex scenes. The animation motion is more natural, the user operation feedback is more accurate, and the animation production efficiency and interactive experience are significantly improved.
Smart Images
Figure CN121767515A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of skeletal display technology, and more particularly to an interactive method for displaying skeletal animation. Background Technology
[0002] Skeletal animation is a widely used animation technique in the field of computer graphics. Its core logic lies in simulating the hierarchical structure of a biological skeletal system, thereby realizing the construction and motion deformation control of 3D models. It associates and binds the geometric surface of the model with the virtual skeletal system, allowing the rotation, swinging and other movements of the bones to directly drive the skin to produce corresponding deformations, ultimately presenting a natural and smooth complex dynamic effect. This technique, by controlling the movement of the model, can realistically reproduce the limb movements and posture changes of the character.
[0003] Traditional methods, which rely solely on physical motion capture data, can ensure the continuity of fast-moving animations, but users cannot easily fine-tune details. When relying solely on manual data manipulation by users, adjustments can be made, but this can easily lead to stuttering and stiff transitions in fast-moving animations. Neither of these methods can easily meet the animation production and interaction needs in complex scenarios. Therefore, it is necessary to propose an interactive display method for skeletal animation. Summary of the Invention
[0004] The purpose of this invention is to address the shortcomings of existing technologies, which, while ensuring the continuity of fast-paced movements, make it difficult for users to fine-tune details. When relying solely on manual data manipulation by the user, adjustments can be made, but this can easily lead to stuttering and stiff transitions in fast-paced movements. Neither of these two modes can easily meet the animation production and interaction needs in complex scenarios. Therefore, this invention proposes an interactive display method for skeletal animation.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for interactive display of skeletal animation, characterized in that the method is applied to an interactive display terminal, and the method includes: The interactive display terminal generates a standardized hierarchical joint tree based on multi-view images or line drawings. It uses dynamic isolation of conflict areas and weight self-optimization algorithm to allocate skin weights, builds a parameterized linkage rule library for joint motion, and forms an adjustable skeletal skin basic structure. The interactive display terminal integrates physical motion capture data, text command parsing data, and user operation data. The joint tree is synchronously driven by a multi-source data spatiotemporal collaborative fusion engine, and the motion smoothness is optimized by scene-adaptive dynamic weight allocation. The interactive display terminal analyzes the user's operation trajectory based on the user behavior intent recognition model, generates personalized parameter templates, builds a multimodal interactive closed loop, and adapts the joint tree movement in combination with an intelligent collision linkage response mechanism. The interactive display terminal supports simultaneous editing by multiple users through a cloud-based collaborative platform, collects interactive performance data, and reverse-optimizes the basic skeletal skin structure and the parameterized linkage rules to form a closed-loop process.
[0006] The above technical solution further includes: The joint tree is a collection of joints constructed according to the specifications of human and animal skeletal structures based on multi-view images or line drawings. There is a parent-child relationship between the joints. Starting from the root joint, it branches downward to form the trunk, limbs, and head sub-joint hierarchy. The movement of the sub-joints is constrained by the parent joint. The overall structure can support skin weight allocation and linkage motion control driven by multi-source data, providing a basic skeleton for the regulation and smooth display of skeletal animation.
[0007] The method of assigning skinning weights using a dynamic isolation and self-optimization algorithm for conflict regions includes: Identify the conflict regions in the standardized hierarchical joint tree that are prone to motion interference, and delineate the boundaries of the regions according to the range of influence of adjacent joint movements; Initialize the skin weight allocation matrix and dynamically adjust the weight values based on the joint motion amplitude; The overlap of weights in the conflict areas is reduced through iterative calculations to ensure the fit between the skin and the bone movement.
[0008] The conflict region refers to the area in the standardized hierarchical joint tree where the range of motion of adjacent joints is likely to overlap or where the joints are likely to interfere with each other during movement. The formation of this region is directly related to the parent-child relationship between joints, the range of motion, and the skin coverage.
[0009] The integrated physical motion capture data, text command parsing data, and user operation data include: The text instruction parsing extracts action keywords through semantic word segmentation, and converts them into angle and time sequence parameters of joint movement trajectory through a preset action mapping table; The user operation data includes mouse drag coordinates, joint bending angles controlled by gestures, and touch press duration.
[0010] The preset action mapping table is a correspondence table used to associate the action keywords with the joint motion trajectory parameters. The preset action mapping table contains the action keywords, the corresponding angle of the joint motion trajectory, and the core information of the time series parameters. During the text instruction parsing process, the key parameters of the joint movement can be directly obtained by quickly querying the table for matching items using the extracted action keywords. This provides a standardized instruction conversion basis for the multi-source data to collaboratively drive the joint tree, ensuring that the text instructions can be converted into motion control signals for the skeletal animation.
[0011] The scene adaptive dynamic weight allocation includes: In fast-moving scenarios, the physical motion capture data is prioritized and assigned a higher weight to ensure motion continuity; In scenarios requiring fine-tuning, the weighting of user operation data is increased to ensure the accuracy of the adjustments; Simultaneously calculate joint motion inertia compensation values to optimize motion transition effects in different scenarios.
[0012] The user behavior intent recognition model analyzes the user's operation trajectory, extracts core parameters such as joint adjustment frequency, adjustment amplitude, and dwell time, and generates multiple user operation preference labels through clustering algorithm. The model matches action parameter templates according to the labels, and the template parameters are updated in real time with the subsequent user operation trajectory. When constructing the multimodal interaction closed loop, the voice triggering module recognizes voice commands and converts them into joint motion control signals, the tactile feedback module matches the vibration intensity according to the joint bending angle and movement speed, and the motion response module ensures that the response logic of the voice commands, the user operations, and the joint tree motion is consistent.
[0013] The intelligent collision linkage response mechanism includes: using a collision detection algorithm to monitor the distance between the joint tree and virtual scene objects in real time, and triggering a collision warning when the distance is less than a preset safety threshold; Based on the collision location and force, associated actions are generated. These associated actions are consistent with the current motion logic of the joint tree to avoid conflicts, and are synchronously fed back to the haptic module to enhance the interactive immersion.
[0014] The cloud-based collaborative platform supports simultaneous editing by multiple users, sets multi-level permissions, synchronizes editing data in real time, and handles conflicting edits according to preset rules while retaining editing records. When reverse-optimizing the basic skeletal skin structure and the parameterized linkage rules, the interaction performance data is collected, the joint linkage coefficients and weight ratios are adjusted based on the data, and the parameters are updated at fixed intervals.
[0015] The present invention has the following beneficial effects: In this invention, the joint tree is driven by a combination of spatiotemporal collaborative fusion of multi-source data and scene-adaptive weight allocation, which breaks through the limitations of traditional single data-driven methods. It achieves spatiotemporal synchronization of three types of data: physical motion capture, text command parsing, and user operation. Then, the weights are dynamically adjusted according to the scene. Fast action scenes prioritize continuity, while fine-tuned scenes focus on accuracy.
[0016] This invention can significantly improve the smoothness of skeletal animation and the accuracy of interaction, solve the problem of balancing speed and accuracy in animation-driven processes in complex scenes, make animation movements more natural, provide more accurate user feedback, and significantly improve animation production efficiency and interactive experience. Attached Figure Description
[0017] Figure 1 This is a schematic diagram illustrating the workflow of the interactive skeletal animation display method proposed in this invention; Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example
[0019] like Figure 1 As shown, the skeletal animation interactive display method proposed in this invention is applied to an interactive display terminal, and the method includes: The interactive display terminal generates a standardized hierarchical joint tree based on multi-view images or line drawings. It uses dynamic isolation of conflict areas and weight self-optimization algorithm to allocate skin weights, builds a parameterized linkage rule library for joint motion, and forms an adjustable skeletal skin basic structure. The interactive display terminal integrates physical motion capture data, text command parsing data, and user operation data. It synchronously drives the joint tree through a multi-source data spatiotemporal collaborative fusion engine and optimizes motion smoothness by scene-adaptive dynamic weight allocation. The interactive display terminal analyzes user operation trajectories based on user behavior intent recognition models, generates personalized parameter templates, builds a multimodal interactive closed loop, and adapts joint tree movement in combination with an intelligent collision linkage response mechanism. The interactive display terminal supports simultaneous editing by multiple users through a cloud-based collaborative platform, collects interactive performance data, and reverse-optimizes the basic structure of the skeleton skin and parametric linkage rules to form a closed loop throughout the entire process.
[0020] Furthermore, the interactive display terminal first generates a standardized hierarchical joint tree with a clear parent-child relationship level based on multi-view images or line drawings according to general skeletal specifications. It then uses dynamic isolation of conflict areas and weight self-optimization algorithms to allocate skinning weights and constructs a parameterized linkage rule library for joint motion to form an adjustable skeletal skinning basic structure. Subsequently, it integrates physical motion capture data, text command parsing data, and user operation data, and completes data spatiotemporal synchronization through a multi-source data spatiotemporal collaborative fusion engine to drive the joint tree. At the same time, it uses scene-adaptive dynamic weight allocation to optimize motion smoothness. Next, it analyzes user operation trajectories based on user behavior intent recognition models, generates personalized parameter templates, builds a multimodal interaction closed loop, and adapts the joint tree motion with an intelligent collision linkage response mechanism. Finally, it supports multi-person synchronous editing through a cloud-based collaborative platform, collects interaction performance data in real time, and uses this data to reverse-optimize the skeletal skinning basic structure and parameterized linkage rules, ultimately forming a closed-loop workflow of construction, driving, interaction, and optimization.
[0021] The joint tree is a collection of joints constructed according to the norms of human and animal skeletal structures based on multi-view images or line drawings. There is a parent-child relationship between the joints. Starting from the root joint, it branches downward to form the sub-joint hierarchy of the trunk, limbs, and head. The movement of the sub-joints is constrained by the parent joint. The overall structure can support skin weight allocation and linkage motion control driven by multi-source data, providing a basic skeleton for the regulation and smooth display of skeletal animation.
[0022] Furthermore, in the character skeletal animation production scenario, the interactive display terminal first acquires multi-view images or design line drawings of the target character, and then constructs a set of joints composed of multiple joints according to the structural specifications of human or animal skeletons. The parent-child relationship between joints is clarified, and branches downward from the root joint to form the sub-joint hierarchy corresponding to the torso, limbs, and head in sequence. The movement of all sub-joints is constrained by their corresponding parent joints. The final overall structure can effectively support the subsequent skinning weight allocation and joint linkage motion control driven by multi-source data, providing a basic skeleton for the precise control and smooth display of skeletal animation.
[0023] The skin weights are assigned using a dynamic isolation of conflict regions and a self-optimizing weight algorithm, including: Identify conflict areas in the standardized hierarchical joint tree that are prone to motion interference, and delineate the boundaries of these areas according to the range of influence of adjacent joint movements; Initialize the skin weight allocation matrix and dynamically adjust the weight values based on the joint motion amplitude; Iterative calculations reduce the overlap of weights in conflict areas, ensuring the fit between the skin and the skeleton.
[0024] Furthermore, in the production of humanoid character skeletal animation, the interactive display terminal first analyzes the constructed standardized hierarchical joint tree to identify conflict areas prone to motion interference, such as the shoulder joint and upper arm joint, and the hip joint and thigh joint. The boundaries of these areas are then clearly defined according to the respective motion influence range of adjacent joints. Subsequently, the skin weight allocation matrix is initialized, and the skin weight values corresponding to each joint in the matrix are set as the initial baseline values. The corresponding weights are then dynamically adjusted according to the actual motion amplitude of different joints. Finally, the weight allocation is continuously optimized through multiple rounds of iterative calculations to continuously reduce the weight overlap in conflict areas, ultimately ensuring that the character skin maintains a good fit during skeletal movement and avoiding stretching or distortion.
[0025] The conflict region refers to the area in the standardized hierarchical joint tree where the range of motion of adjacent joints is likely to overlap or interfere with each other during movement. The formation of this region is directly related to the parent-child relationship between joints, the range of motion, and the skin coverage.
[0026] Furthermore, in the production of humanoid character skeletal animation, the interactive display terminal, based on the constructed standardized hierarchical joint tree, focuses on analyzing adjacent joints such as the elbow joint and forearm joint, and the knee joint and lower leg joint. Because these joints have clear parent-child relationships and their respective ranges of motion are large, and the skin coverage around the joints is interconnected, the range of motion of the two tends to overlap during joint bending, rotation and other movements, and they will also interfere with each other during movement, ultimately forming these specific areas.
[0027] It integrates physical motion capture data, text command parsing data, and user operation data, including: Text instruction parsing extracts action keywords through semantic word segmentation, and then converts them into angle and time sequence parameters of joint motion trajectories through a preset action mapping table; User operation data includes mouse drag coordinates, joint bending angles controlled by gestures, and touch press duration.
[0028] Furthermore, in the humanoid character skeletal animation interactive scenario, the interactive display terminal first parses the received text instruction "slowly raise arm for 3 seconds," extracts action keywords such as "raise arm," "slowly," and "3 seconds" through semantic word segmentation, and then calls a preset action mapping table to convert these keywords into specific angle parameters and corresponding time series parameters of the shoulder and elbow joint movement trajectories. At the same time, it collects in real time the coordinate data generated by the user dragging the character's arm with the mouse, the bending angle data of the wrist joint under gesture control, and the pressing duration data when adjusting the action by touching the screen. This data is integrated with the physical motion capture data of the actor's real arm raising action acquired simultaneously to complete the collection and preprocessing of the three types of data.
[0029] The preset motion mapping table is a table used to associate motion keywords with joint motion trajectory parameters. The preset motion mapping table contains core information such as motion keywords, corresponding joint motion trajectory angles, and time series parameters. During the text instruction parsing process, the extracted action keywords can be used to quickly look up matching items in the table and directly obtain key parameters of joint movement. This provides a standardized basis for instruction conversion for multi-source data collaboratively driving the joint tree, ensuring that text instructions can be converted into motion control signals for skeletal animation.
[0030] Furthermore, in the humanoid character skeletal animation interaction scenario, the pre-built preset action mapping table stores action keywords such as "waving," "bending," and "turning around," as well as the specific angle parameters of the joint movement trajectory of the shoulder joint, elbow joint, etc., corresponding to each keyword, and time series parameters such as "complete in 2 seconds" and "uniform movement." When the interactive display terminal receives and parses the text instruction "wave quickly for 2 seconds," it first extracts the core action keywords "wave," "quickly," and "2 seconds" through semantic word segmentation, and then quickly queries the matching items in the preset action mapping table based on these keywords to directly obtain the corresponding joint movement angles and time series key parameters. These parameters, as the core component of the text instruction parsing data, work in conjunction with physical motion capture data and user operation data to provide a standardized instruction conversion basis for multi-source data-driven joint trees, ensuring that the text instruction is accurately converted into skeletal animation motion control signals that drive the character to complete the quick waving action.
[0031] Scene-adaptive dynamic weight allocation includes: In fast-moving scenarios, physical motion capture data is prioritized and assigned higher weights to ensure motion continuity; In scenarios requiring fine-tuning, the weighting of user operation data is increased to ensure the accuracy of adjustments; Simultaneously calculate joint motion inertia compensation values to optimize motion transition effects in different scenarios.
[0032] Furthermore, in the interactive production scenario of character skeletal animation, when performing a fast-paced action display of "character running quickly," the interactive display terminal prioritizes responding to the physical motion capture data of the actor's actual running and assigns it a higher weight to ensure the smooth and continuous running action. When switching to the fine adjustment scenario of "adjusting the character's finger grip posture," the terminal automatically increases the weight of the operation data generated by the user through gesture control to ensure the accuracy of finger joint adjustment. At the same time, during the switching between the two scenarios and the connection of actions in the same scenario, the terminal synchronously calculates the motion inertia compensation value of each joint, effectively optimizing the transition effect when starting and stopping actions and switching scenarios, and avoiding stuttering or stiffness.
[0033] The user behavior intent recognition model analyzes the user's operation trajectory, extracts core parameters such as joint adjustment frequency, adjustment amplitude, and dwell time, and generates multiple user operation preference labels through clustering algorithm. The action parameter template is matched according to the label, and the template parameters are updated in real time with the subsequent user operation trajectory. When building a multimodal interactive closed loop, the voice triggering module recognizes voice commands and converts them into joint motion control signals, the haptic feedback module matches the vibration intensity according to the joint bending angle and movement speed, and the motion response module ensures that the response logic of voice commands, user operations and joint tree movements is consistent.
[0034] Furthermore, in the scenario of fine-tuning character skeletal animation, the user behavior intent recognition model of the interactive display terminal analyzes in real time the operation trajectory of the user repeatedly dragging the character's elbow joint and fine-tuning the wrist joint. It extracts the core parameters of joint adjustment frequency, amplitude of each adjustment, and duration of stay at a specific angle during this process. These parameters are classified through a clustering algorithm to generate multiple user operation preference tags such as "fine-tuning type" and "high-frequency switching type". The corresponding action parameter template is retrieved according to the matched "fine-tuning type" tag, and the template parameters are updated in real time as the user continues to adjust the angle of the finger joints. At the same time, a multimodal interaction closed loop is built. The voice trigger module accurately recognizes the user's voice command "bend the elbow 30 degrees" and converts it into the corresponding joint motion control signal. The haptic feedback module matches the corresponding vibration intensity according to the actual bending angle and movement speed of the elbow joint (the larger the angle and the faster the speed, the more obvious the vibration). The motion response module ensures that the voice command, the user's synchronous mouse drag operation, and the motion response logic of the joint tree are consistent, realizing a smooth connection of multimodal interaction.
[0035] The intelligent collision linkage response mechanism includes: using a collision detection algorithm to monitor the distance between the joint tree and virtual scene objects in real time, and triggering a collision warning when the distance is less than a preset safety threshold; Based on the collision location and force, related actions are generated. These related actions are consistent with the current motion logic of the joint tree to avoid conflicts, and are simultaneously fed back to the haptic module to enhance the interactive immersion.
[0036] Furthermore, in the virtual scene interaction of character skeletal animation, the interactive display terminal monitors the distance between the character's joint tree (such as arm joints and leg joints) and objects such as tables and walls in the virtual scene in real time through a collision detection algorithm. When the character's arm joint approaches the virtual table and the distance is less than a preset safety threshold, a collision warning is immediately triggered. Then, based on the collision location (arm joint) and the estimated collision force, a related action of "the arm naturally retracts and slightly lifts" is generated. This related action is consistent with the character's current "walking forward" motion logic to avoid motion conflicts. At the same time, the collision feedback signal is synchronized to the haptic module so that the user can feel the corresponding intensity of vibration when operating the terminal, further enhancing the interactive immersion.
[0037] The cloud-based collaborative platform supports simultaneous editing by multiple users, allows setting multi-level permissions, synchronizes editing data in real time, and handles conflicting edits according to preset rules while retaining editing records. When reverse-optimizing the basic structure of the skeleton skin and the parameterized linkage rules, interaction performance data is collected, and the joint linkage coefficients and weight ratios are adjusted based on the data, and the parameters are updated at fixed intervals.
[0038] Furthermore, in scenarios involving multi-team collaboration in creating character skeletal animations, the cloud-based collaborative platform sets up multi-level permissions for project members, including administrators, editors, and viewers. Administrators can modify core parameters, editors can adjust animation movements, and viewers can only browse progress. The platform synchronizes all members' editing data in real time. When two editors simultaneously modify the same joint motion parameters and a conflict arises, the platform handles the situation according to the preset rule of "the later editor's confirmation overwrites the previous edited content," and retains a record of every editing operation for retrospective purposes. At the same time, the interactive display terminal continuously collects interactive performance data such as user operation response latency, motion smoothness scores, and collision conflict rates. Based on this data analysis, the technical team optimizes the direction, specifically adjusting the joint linkage coefficients of the skeletal skinning basic structure and the weighting of parameterized linkage rules. The parameters are updated and optimized on a fixed weekly basis to continuously improve the animation display effect and interactive experience.
[0039] In this embodiment, the specific implementation method is as follows: in the multi-team collaborative interactive display scenario of humanoid character skeletal animation, the interactive display terminal first acquires multi-view design images and line drawings of the target character, constructs a standardized hierarchical joint tree composed of multiple joints according to the humanoid skeletal structure specifications, clarifies the parent-child relationship between joints, and branches downward from the root joint to form the trunk, limbs, and head sub-joint hierarchy, with the movement of sub-joints constrained by the parent joint. Subsequently, the joint tree is analyzed to identify conflict areas such as the shoulder joint and upper arm joint, and the knee joint and lower leg joint, which are prone to interference due to large movement amplitude and skin coverage connection. Boundaries are defined according to the influence range of adjacent joint movement, the skin weight allocation matrix is initialized and the weight value is dynamically adjusted according to the joint movement amplitude. The overlap of weights in conflict areas is reduced through multiple rounds of iterative calculation. At the same time, a joint movement parameterized linkage rule library containing parameters such as joint rotation angle and movement speed is constructed to form a skeletal skin basic structure that can be precisely controlled. Next, the terminal integrates physical motion capture data of the actor's real movements, semantic segmentation to extract keywords from text commands such as "wave quickly for 2 seconds", and motion parameters converted by a preset motion mapping table (storing keywords such as "wave" and "bend over" and corresponding joint angles and time series parameters), as well as operation data such as user mouse drag coordinates, joint bending angles controlled by gestures, and touch press duration. Through a multi-source data spatiotemporal collaborative fusion engine, the three types of data are spatiotemporally synchronized by timestamp alignment. Then, a scene-adaptive dynamic weight allocation strategy is adopted. In fast action scenarios such as "character running quickly", physical motion capture data is given priority and higher weight is allocated to ensure continuity. In fine adjustment scenarios such as "adjusting finger grip posture", the weight ratio of user operation data is increased to ensure accuracy. At the same time, the joint motion inertia compensation value is calculated to optimize the transition effect and synchronously drive the joint tree movement. Subsequently, the terminal uses a user behavior intent recognition model to analyze in real time the user's operation trajectory of repeatedly dragging the elbow joint and fine-tuning the wrist joint, extracting core parameters such as joint adjustment frequency, amplitude and dwell time. Through clustering algorithm, user operation preference labels such as "fine-tuning type" are generated and matched with corresponding action parameter templates. The template parameters are updated in real time with subsequent operation trajectories. At the same time, a multimodal interaction closed loop is built. The voice trigger module recognizes the voice command "bend the elbow 30 degrees" and converts it into a control signal. The tactile feedback module matches the vibration intensity according to the joint bending angle and movement speed. The action response module ensures that the voice command, user operation and joint tree movement logic are consistent. With the intelligent collision linkage response mechanism, the collision detection algorithm monitors the distance between the joint tree and objects such as tables and walls in the virtual scene in real time. When the distance is less than the preset safety threshold, a collision warning is triggered. Based on the collision part and force, an associated action consistent with the current movement logic (such as the arm naturally retracting) is generated and synchronously fed back to the tactile module. Finally, a cloud-based collaborative platform was used to set up multi-level permissions for project members, including administrators, editors, and viewers. Editing data was synchronized in real time. When conflicting edits occurred, the "later editor confirms overwrite" rule was followed, and the entire editing record was retained. The terminal continuously collected interactive performance data such as user operation response latency, motion smoothness score, and collision conflict rate. Based on this data analysis, the technical team optimized the direction and made targeted adjustments to the joint linkage coefficients of the basic skeletal skin structure and the weight ratio of parameterized linkage rules. The optimization parameters were updated and optimized on a fixed weekly basis, ultimately forming a closed loop of the entire process of construction, driving, interaction, and optimization, achieving efficient production, smooth display, and continuous iteration of skeletal animation.
[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their likenesses.
Claims
1. A method for skeletal animation interactive presentation, characterized in that, The method is applied to an interactive display terminal, and the method comprises: The interactive display terminal generates a standardized hierarchical joint tree based on multi-view images or sketches, adopts a conflict area dynamic isolation and weight self-optimization algorithm to distribute skin weight, constructs a joint motion parameterization linkage rule library, and forms a controllable bone skin basic structure; The interactive display terminal integrates physical motion capture data, text instruction analysis data and user operation data, synchronously drives the joint tree through a multi-source data space-time collaborative fusion engine, and adopts scene adaptive dynamic weight distribution to optimize motion fluency; The interactive display terminal analyzes the user operation trajectory based on a user behavior intention recognition model, generates a personalized parameter template, builds a multi-modal interactive closed loop, and adapts the joint tree motion in combination with an intelligent collision linkage response mechanism; The interactive display terminal supports multi-person synchronous editing through a cloud collaborative platform, collects interactive efficiency data, reversely optimizes the bone skin basic structure and the parameterization linkage rule, and forms a full-process closed loop.
2. The skeletal animation interactive presentation method of claim 1, wherein, The joint tree is a joint set constructed based on multi-view images or sketches according to the specification of human or animal skeletal structure; The joints have a parent-child association relationship, and a trunk, limbs and head sub-joint level is formed from the root joint as a starting point, the motion of a sub-joint is constrained by a parent joint, and the overall structure can support skin weight distribution, linkage motion control under multi-source data driving, and provides a basic skeleton for the regulation and smooth display of bone animation.
3. The skeletal animation interactive presentation method of claim 1, wherein, The conflict area dynamic isolation and weight self-optimization algorithm for distributing skin weight comprises: Identifying the conflict area in the standardized hierarchical joint tree which is prone to motion interference, and delimiting the area boundary according to the influence range of adjacent joint motion; Initializing a skin weight distribution matrix, and dynamically adjusting the weight value based on joint motion amplitude; Reducing the weight overlap degree of the conflict area through iterative calculation to ensure the fit degree of skin and bone motion.
4. The skeletal animation interactive presentation method of claim 3, wherein, The conflict area refers to an area in the standardized hierarchical joint tree in which the adjacent joints are prone to overlap in the motion process or are prone to mutual interference when the joints move, and the formation of the area is directly related to the parent-child association relationship between joints, motion amplitude and skin coverage range.
5. The skeletal animation interactive presentation method of claim 1, wherein, The integration of physical motion capture data, text instruction analysis data and user operation data comprises: The text instruction analysis extracts action keywords through semantic word segmentation, and converts the angle and time sequence parameters of joint motion trajectory into a preset action mapping table; The user operation data includes mouse dragging coordinates, joint bending angles controlled by gestures and touch pressing time length.
6. The skeletal animation interactive presentation method of claim 5, wherein, The preset action mapping table is a correspondence table for associating the action keywords with the joint motion trajectory parameters, and the preset action mapping table contains the action keywords, corresponding angle and time sequence parameter core information of the joint motion trajectory. In the text instruction analysis process, the key parameters of the joint motion can be directly obtained by matching the extracted action keywords in the quick query table, providing a standardized instruction conversion basis for the multi-source data collaborative driving of the joint tree, and ensuring that the text instruction can be converted into the motion control signal of the skeletal animation.
7. The skeletal animation interactive presentation method of claim 1, wherein, The scene adaptive dynamic weight distribution includes: In the fast action scene, the physical motion capture data is preferentially responded and a higher weight is allocated to ensure the continuity of the motion; In the fine adjustment scene, the weight proportion of the user operation data is increased to ensure the adjustment accuracy; The joint motion inertia compensation value is calculated synchronously to optimize the motion transition effect in different scenes.
8. The skeletal animation interactive presentation method of claim 1, wherein, The user behavior intention recognition model analyzes the user operation trajectory, extracts the core parameters of joint adjustment frequency, adjustment amplitude, and stay duration, generates multiple user operation preference labels through clustering algorithm classification, matches the action parameter template according to the label, and updates the template parameters in real time according to the subsequent user operation trajectory; When building the multi-modal interaction closed loop, the voice trigger module identifies the voice instruction and converts it into a joint motion control signal, the haptic feedback module matches the vibration intensity according to the joint bending angle and motion speed, and the action response module ensures that the response logic of the voice instruction, the user operation, and the joint tree motion is consistent.
9. The skeletal animation interactive presentation method of claim 1, wherein, The intelligent collision linkage response mechanism includes: using a collision detection algorithm to monitor the distance between the joint tree and virtual scene objects in real time, and triggering a collision warning when the distance is less than a preset safety threshold; Based on the collision site and intensity, the associated action is generated, which is consistent with the current motion logic of the joint tree to avoid conflicts, and is fed back to the haptic module to enhance the interactive immersion.
10. The skeletal animation interactive presentation method of claim 1, wherein, The cloud collaborative platform supports multi-person synchronous editing, sets multiple levels of permissions, synchronizes the editing data in real time, and processes the conflict editing according to the preset rules and retains the editing records; When the skeletal skinning basic structure and the parameterized linkage rules are optimized in reverse, the interaction efficiency data is collected, the joint linkage coefficient and weight proportion are adjusted based on the data, and the parameters are updated at a fixed period.