Vehicle Localization Across Outdoor-Indoor Map Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle localization and mapping systems face challenges in accurately correcting the estimated location of vehicles based on loop closure detection, particularly when transitioning from outdoor to indoor environments, especially when GPS signals are interrupted.
Innovation Solution
A vehicle control apparatus that includes a camera and sensors to generate a point cloud for indoor environment mapping, using simultaneous localization and mapping (SLAM) to correct the vehicle's location by integrating it with an external environment map, thereby determining the start and end points of the indoor environment map.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS signals are used for vehicle localization, then location estimation accuracy is improved in outdoor environments, but the system becomes unreliable when transitioning to indoor environments where GPS signals are interrupted
Solution Approach 1:
The patent divides the environment into distinct segments (outdoor and indoor) and applies different localization methods appropriate to each segment. GPS is used for outdoor localization while SLAM with camera-based mapping is used for indoor localization, allowing the system to maintain reliability across different environments.
Solution Approach 2:
The patent introduces an intermediary mechanism (SLAM system with camera and point cloud processing) that bridges the gap between GPS-based outdoor localization and indoor localization requirements. This intermediary system enables seamless transition and maintains continuous location estimation reliability.
2Adaptability or versatility
If SLAM is used for indoor environment mapping, then GPS dependency is reduced, but location correction accuracy deteriorates without loop closure detection
Solution Approach 1:
The patent implements feedback through loop closure detection, where the system continuously monitors for recognizable environmental features and corrects accumulated drift in SLAM-based location estimates. This feedback mechanism maintains measurement precision by detecting and correcting errors in real-time.
Solution Approach 2:
The patent performs preliminary mapping actions by creating and storing environmental maps during initial exploration phases. These pre-established maps serve as reference data for subsequent loop closure detection and location correction, improving precision without requiring real-time complex processing.
3Ease of operation
If comprehensive environment mapping is performed, then navigation capability is improved, but system complexity increases
Solution Approach 1:
The patent applies local quality by creating detailed environment maps only in the specific indoor areas where the vehicle operates, rather than attempting to map entire buildings or environments. This localized mapping approach maintains navigation capability while reducing system complexity and computational requirements.
Solution Approach 2:
The patent performs partial mapping actions by focusing on mapping only the necessary pathways and relevant environmental features required for navigation, rather than comprehensively mapping all areas. This selective approach improves ease of operation while keeping device complexity manageable.
Data Source
AI summary
An apparatus, including a camera, a processor, and a memory, is configured to identify that a vehicle enters a point in an indoor environment from a point in an external environment, obtain a point cloud for at least one object identified using the camera based on the vehicle entering the point in the indoor environment, generate an indoor environment map representing at least a portion of the indoor environment along a movement path of the vehicle in the indoor environment by using the point cloud, and determine the point in the indoor environment as a start point of the indoor environment map by mapping the point of the external environment and the point of the indoor environment using an external environment map representing the external environment.