Mixed Reality Driving Images for Virtual Object Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity and diversification of autonomous driving vehicle parts and software pose challenges in efficiently testing these vehicles, leading to place limitations and increased time and cost for performance evaluation in real driving environments.
Innovation Solution
A method and system for generating a mixed reality image by combining real and virtual images using cameras with identical external and internal parameters, allowing for the creation of a mixed reality image that simulates the driving environment, enabling virtual object testing without physical objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real driving environments are used for testing autonomous driving vehicles, then testing reliability is improved, but place limitations and increased time and cost occur
Solution Approach 1:
The patent creates virtual copies of driving environments, vehicles, and objects using simulation models. These virtual replicas enable comprehensive testing without requiring actual physical presence, thereby eliminating place limitations and reducing testing time while maintaining reliability through realistic scenario reproduction
Solution Approach 2:
The system performs preliminary generation of virtual driving environments and objects before actual testing begins. By pre-configuring test scenarios, road layouts, and object positions in the virtual environment, the system eliminates the need for time-consuming setup in real locations, significantly reducing overall testing time
2Reliability
If real driving environments are used for testing autonomous driving vehicles, then testing realism is improved, but place limitations occur
Solution Approach 1:
The virtual environment system serves multiple testing functions simultaneously - it can simulate various road types, weather conditions, traffic scenarios, and object configurations within a single platform. This universal testing environment eliminates place limitations while maintaining realism, allowing diverse scenario testing without requiring multiple physical locations
Solution Approach 2:
The patent transitions from physical three-dimensional space to virtual computational space, adding the dimension of programmable scenario control. This enables unlimited adaptability in creating and modifying test scenarios without being constrained by physical geography, while maintaining realistic visual and physical properties through advanced rendering and physics engines
3Reliability
If more virtual objects are added to simulate complex driving environments, then testing comprehensiveness is improved, but image processing complexity increases
Solution Approach 1:
The system segments the complex virtual environment into distinct components - road geometry, static objects, dynamic vehicles, pedestrians, and environmental conditions. Each segment is independently modeled and rendered, allowing comprehensive scenario construction while managing processing complexity through modular organization and selective refinement of different environment elements
Data Source
AI summary
A method for generating a mixed reality image related to a driving environment is performed by one or more processors, and includes acquiring a real image related to the driving environment, generating a first virtual image including an object of the driving environment, and generating a mixed reality image including the object of the driving environment based on the real image and the first virtual image.