Localized 3D-to-2D Map Projection for Autonomous Driving Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving systems face navigation challenges due to distortions in 2D maps, which are less accurate than 3D maps but more computationally efficient, particularly when projecting spherical Earth surfaces to 2D surfaces, leading to inaccuracies in geometric calculations.
Innovation Solution
A control system for autonomous vehicles that shifts and projects 3D maps to 2D maps using techniques like Mercator, UTM, Equirectangular, Cassini, cylindrical, and pseudocylindrical projections, selecting a base point to minimize distortions within an effective area, allowing for efficient navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If 2D maps are used for navigation, then computational efficiency is improved, but location accuracy deteriorates due to spherical to planar projection distortions
Solution Approach 1:
The patent applies local quality by creating a localized 2D map projection centered on the vehicle's current position rather than using a global projection. This ensures that the area immediately surrounding the vehicle maintains high geometric accuracy while still benefiting from 2D computational efficiency. The projection parameters are specifically tuned for the local region, minimizing distortion where it matters most for navigation decisions.
Solution Approach 2:
The patent dynamically adjusts projection parameters based on the vehicle's position, speed, and intended route. By changing the projection center, scale, and orientation parameters adaptively, the system maintains accurate geometric representations in the relevant navigation area while preserving the computational advantages of 2D maps. This allows the system to optimize for both accuracy and efficiency in real-time.
2Measurement precision
If 3D maps are used for navigation, then location accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent extracts only the essential 3D geometric information needed for navigation and represents it in a simplified 2D format. Rather than processing full 3D spatial data structures, the system extracts position, bearing, and distance information and represents it in a 2D map with adjusted projection parameters. This extraction approach maintains the accuracy benefits of 3D mapping while eliminating the computational burden of processing complete 3D data.
3Ease of operation
If standard 2D projection techniques are used, then ease of operation is improved, but geometric calculation accuracy deteriorates
Solution Approach 1:
The patent modifies standard 2D projection parameters to optimize for geometric calculation accuracy. Specifically, it adjusts the projection center to the vehicle's current position, sets the appropriate scale factor for the operating area, and configures the orientation to align with the vehicle's heading or intended route. These parameter changes maintain the simplicity and ease of operation of standard 2D projections while significantly improving the accuracy of distance, bearing, and intersection calculations.
Data Source
AI summary
Systems and methods for projecting a three-dimensional (3D) surface to a two-dimensional (2D) surface for use in autonomous driving are disclosed. In one aspect, a control system for an autonomous vehicle includes a processor and a computer-readable memory in communication with the processor and having stored thereon computer-executable instructions to cause the processor to: receive a 3D map including a plurality of objects, determine a base point in the 3D map, shift the objects in the 3D map based on the base point, project the objects in the shifted 3D map to a 2D map, and output the 2D map.