Virtual Scene Reconstruction Knowledge Graph for Semantic Scene Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual reality systems struggle to efficiently extract and reconstruct specific scenes with a minimum and proper time range in a minimum and proper space, as they lack semantic fragmentation of scene constituting elements, making it difficult to select scenes based on semantic information such as events or actions.
Innovation Solution
A scene recording and reconstructing device that records geometry, record, and context information as a knowledge graph, allowing for the reconstruction of desired scenes by identifying and combining relevant information from the graph.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complete scene information is recorded and reconstructed in virtual space, then scene reconstruction capability is improved, but the ability to extract and select specific scenes based on semantic information deteriorates
Solution Approach 1:
The patent segments scene information into three distinct types: geometry information (spatial structure), record information (temporal events), and context information (semantic relationships). This segmentation allows the system to maintain complete scene data for reliable reconstruction while enabling selective extraction of specific semantic components through the knowledge graph, resolving the contradiction between complete reconstruction capability and ease of scene selection.
2Adaptability or versatility
If all recorded scenes are stored for future reference, then scene availability is improved, but the time required to search and identify specific scenes increases
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary structure that indexes and organizes scene information semantically. The knowledge graph acts as a mediator between the stored scenes and user queries, enabling efficient search and identification of specific scenes based on semantic relationships, events, or actions without requiring time-consuming sequential review of all recorded scenes.
3Quantity of substance
If scene information is recorded without semantic fragmentation, then data completeness is improved, but the ability to reconstruct scenes with proper time and space ranges deteriorates
Solution Approach 1:
The patent segments scene data into three distinct categories (geometry, record, context) and represents their relationships through a knowledge graph. This segmentation maintains data completeness while enabling precise reconstruction by allowing the system to selectively combine specific geometry, temporal, and contextual information as needed, achieving both completeness and precision.
Solution Approach 2:
The patent implements dynamic scene reconstruction capability through the knowledge graph, which allows the system to adaptively select and combine scene components based on user requirements. The reconstruction process is dynamic rather than static, enabling adjustment of time ranges and spatial boundaries while maintaining data completeness through the comprehensive knowledge graph structure.
Data Source
AI summary
In order to attain the above object, a scene recording and reconstructing device that records and reconstructs scenes in a virtual space, includes a geometry recording unit that records geometry information which describes a shape or an appearance of an object constituting a scene, a record recording unit that records record information which is time-series information on an event performed by an object, a knowledge graph recording unit that records a context of a scene as a knowledge graph, and a scene reconstruction unit that reconstructs a desired scene in a virtual space by identifying a reconstruction target scene from a context of a scene recorded in the knowledge graph recording unit, acquiring geometry information and record information required to reconstruct the desired scene, and re-combining the geometry information and the record information.