Multi-Level Occupancy Grids for 3D Roadway Occlusion Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current occupancy and occlusion grids in autonomous vehicles are limited to two-dimensional representations, failing to accurately account for multi-level traffic infrastructure such as bridges, overpasses, and tunnels, leading to incorrect identification of occupied or occluded areas.
Innovation Solution
The development of a method to generate a multi-level occupancy and occlusion grid, where each grid cell represents the probability of occlusion and occupancy at each level of a 3D map, using a 3D map that includes ground height information for multiple levels, and associating each cell with a unique index and multiple gridlets corresponding to the number of height levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a two-dimensional occupancy and occlusion grid is used, then the system complexity is low and the grid is easy to process, but the accuracy of representing multi-level traffic infrastructure is insufficient
Solution Approach 1:
The patent transitions from a two-dimensional grid to a three-dimensional grid structure that incorporates vertical height information. Each grid cell includes a height component, enabling the representation of multi-level traffic infrastructure such as bridges, overpasses, and tunnels. This dimensional expansion allows accurate identification of occupied and occluded areas at different elevation levels while maintaining the grid's systematic organization through indexed cells and associated gridlets for each height level
2Loss of information
If a three-dimensional multi-level grid is implemented, then the representation accuracy of multi-level infrastructure is improved, but the data storage requirements and processing complexity increase
Solution Approach 1:
The patent segments the three-dimensional space into discrete grid cells, each associated with multiple gridlets corresponding to different height levels. This segmentation allows the system to represent only the necessary vertical layers where infrastructure exists, rather than storing data for all possible height levels throughout the entire space. The indexed cell structure enables efficient storage and retrieval of occupancy and occlusion information at specific locations and elevations
Data Source
AI summary
This document discloses methods of acquiring a 3D map of an area with multiple heights. The 3D map may include tiles, each of which provides ground height information for one or more of the plurality of levels. A grid corresponding to the 3D map is generated. The grid may divide each of the plurality of tiles into a plurality of cells, The grid may include: (a) a unique index assigned to each of the plurality of cells, in which the unique index is convertible to an index of a tile of the 3D map in which that cell is included, and (b) a plurality of gridlets associated with each of the plurality of cells, wherein a number of the plurality of gridlets associated with a cell is equal to a number of height levels represented by a tile of the 3D map in which that cell is included.