AR Scene Image Overlay Using Touch-Drawn 3D Virtual Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality (AR) technologies for creating personalized scene images are costly, time-consuming, and limited by pre-configured virtual models, which occupy significant storage space and incur high labor costs, while offering limited user customization and interactivity.
Innovation Solution
A method and apparatus for generating personalized scene images by acquiring a touch event trajectory in a screen region, projecting it into a space coordinate system, reconstructing a virtual model based on the device's position and posture, and overlaying real-world images with the model view, allowing users to create customized virtual models directly on their devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If pre-configured virtual models are used for augmented reality, then the implementation is straightforward, but storage space is significantly occupied and labor costs are high
Solution Approach 1:
The patent uses 2D trajectory data as a simplified copy or representation of the 3D virtual model creation process. Instead of storing and processing complete 3D models, the system captures user drawing trajectories on the screen and projects them into 3D space, effectively copying the essential shape information while discarding redundant data, thus reducing storage requirements while maintaining model creation capability
Solution Approach 2:
The patent extracts only the necessary shape information from the virtual model creation process by capturing touch event trajectories. This extraction approach separates the essential geometric data (trajectory points) from the complete model data structure, allowing the system to generate 3D models on-demand from minimal 2D input data, thereby significantly reducing storage space occupation
2Ease of manufacture
If pre-configured virtual models are used for augmented reality, then the implementation is straightforward, but labor costs are high
Solution Approach 1:
The system enables users to create their own virtual models directly through touch gestures on the screen without requiring professional 3D modeling tools or expert intervention. The automatic projection algorithm converts simple 2D drawings into 3D models autonomously, making the model creation process self-service and eliminating the need for expensive professional modeling labor
Solution Approach 2:
The patent replaces the complex mechanical process of professional 3D modeling with an automated algorithmic system. The projection algorithm automatically converts 2D trajectory data into 3D model structures without manual intervention, substituting computational automation for human labor in the model creation process, thereby reducing labor costs
3Ease of manufacture
If pre-configured virtual models are used, then model creation is simple, but model diversity is limited and user customization is restricted
Solution Approach 1:
The system transitions from static pre-configured models to dynamic user-created models. Users can draw different trajectories in real-time to create diverse virtual models on-demand, making the model creation process dynamic and adaptable to various user needs and preferences, thereby significantly increasing model diversity and customization options
Solution Approach 2:
The patent enables users to create 3D virtual models by drawing 2D trajectories on the screen. This dimensionality transformation approach allows simple 2D gestures to generate complex 3D structures, making model creation accessible to ordinary users while maintaining the capability for diverse and customized model designs in three-dimensional space
4Manufacturing precision
If complex 3D modeling is performed in real-time, then model accuracy is high, but processing time increases
Solution Approach 1:
The system performs preliminary action by capturing the touch trajectory data during the user's natural drawing process. The trajectory points are recorded in real-time as the user draws, and the 3D model projection is performed automatically based on this pre-captured data, eliminating the need for subsequent complex modeling operations and reducing overall processing time while maintaining model accuracy
Data Source
AI summary
A personalized scene image processing method is provided for a terminal device. The method includes acquiring, according to a touch event triggered in a screen region of the terminal device, a trajectory of the touch event in the screen region; generating a virtual model, according to a projection of the trajectory of the touch event in a space coordinate system; reconstructing a model view of the virtual model mapped within a field of view of the terminal device, according to a position and posture of the terminal device in the space coordinate system; and overlaying a scene image acquired by the terminal device in the position and posture with the model view to obtain a personalized scene image.