3D Object Model Storage in ISO BMFF with Adaptive LOD Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression techniques struggle to efficiently store and distribute 3D object models, such as user avatars, in extended reality (XR) sessions, particularly due to varying network connections and processing capabilities among users, leading to suboptimal bandwidth and processing resource utilization.
Innovation Solution
Utilizing the ISO Base Media File Format (ISO BMFF) to store 3D object models with multiple levels of detail (LODs), allowing devices to adaptively retrieve and render models based on distance, network bandwidth, and processing capabilities, ensuring efficient distribution and rendering of high-quality 3D objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple LODs of 3D object models are stored in ISO BMFF files, then adaptability to different network connections and processing capabilities is improved, but device complexity increases
Solution Approach 1:
The patent segments the 3D object model into multiple levels of detail (LODs), where each LOD represents a simplified version of the model with varying levels of geometric and textural complexity. This segmentation allows the system to provide different model versions to users based on their network capabilities and device performance, resolving the contradiction between adaptability and complexity by organizing the model data into manageable, progressively detailed layers.
Solution Approach 2:
The patent implements dynamic LOD selection where the system automatically determines which model version to retrieve and render based on real-time assessment of network bandwidth, device processing power, and scene requirements. This dynamic adaptation mechanism allows the system to optimize the balance between model quality and resource consumption without requiring manual configuration, thereby improving adaptability while managing device complexity through automation.
2Productivity
If lower LOD models are retrieved for distant objects, then processing operations and bandwidth requirements are reduced, but rendering quality deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the level of detail for different spatial regions in the scene. Distant objects are rendered at lower LODs to conserve bandwidth and processing resources, while nearby objects are rendered at higher LODs to maintain rendering quality. This spatially adaptive approach resolves the contradiction by recognizing that quality requirements are not uniform across the entire scene, allowing optimization of resources where quality can be compromised without affecting user experience.
Solution Approach 2:
The patent implements partial action by selectively retrieving and rendering only the necessary portions of 3D models based on viewer distance and attention. Instead of always loading complete high-quality models, the system loads only the appropriate LOD level for each object's position in the scene. This partial approach reduces overall processing operations and bandwidth requirements while maintaining sufficient rendering quality for objects that matter most to the user experience.
3Manufacturing precision
If higher LOD models are prioritized for closer objects, then rendering quality is maintained, but processing resources are consumed more heavily
Solution Approach 1:
The patent employs dynamic resource allocation where processing resources are continuously adjusted based on real-time scene analysis. The system monitors object distances, network conditions, and device performance to dynamically switch between different LOD levels. This dynamic strategy ensures that high rendering quality is maintained for nearby objects that require it, while automatically reducing processing resource consumption for distant objects, thereby resolving the contradiction between quality and resource usage through adaptive management.
Data Source
AI summary
An example device for retrieving media data includes a memory for storing media data; and a processing system implemented in circuitry and configured to: retrieve data representing a three-dimensional (3D) object model and one or more levels of detail (LODs) for the 3D object model; send a request to the server device to access data for the 3D object model at one of the LODs, the data for the 3D object model at the one of the LODs including a size, complexity, and components of the 3D object model for the one of the LODs; and receive the data for the 3D object model at the one of the LODs in response to the request, the data for the 3D object model at the one of the LODs having the size, the complexity, and the components of the 3D object model for the one of the LODs.