Dynamic Occupancy Grid Aligned to Lane Curvature in ADAS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional occupancy grids in advanced driver assistance systems (ADAS) require significant computational effort and memory due to the inclusion of numerous unnecessary cells, especially when lanes have strong curvature, leading to inefficiencies in processing and data transfer.
Innovation Solution
A dynamic occupancy grid is generated by detecting the lane course using a detection system and defining cells within a restricted region of interest, reducing the lateral size of the grid and increasing resolution only where necessary, thereby reducing memory usage and computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a conventional occupancy grid with a fixed vehicle coordinate system is used to ensure collision-free space monitoring, then the coverage area is comprehensive, but the number of cells increases significantly leading to high computational effort and memory usage
Solution Approach 1:
The patent transforms the static vehicle coordinate system into a dynamic lane coordinate system that adapts to the actual lane geometry. The coordinate system rotates and translates according to the lane's curvature and orientation, allowing the occupancy grid to dynamically adjust its configuration. This dynamic adaptation enables the system to maintain comprehensive monitoring coverage while reducing the number of cells required, as the grid aligns with the actual road geometry rather than forcing a fixed rectangular overlay.
2Reliability
If the number of cells is increased in lateral direction to guarantee collision-free space, then the safety coverage is improved, but the memory requirements and processing load increase significantly
Solution Approach 1:
The patent implements local quality by allowing different regions of the occupancy grid to have different cell densities and resolutions. Areas closer to the host vehicle and within the actual lane boundaries use finer cell resolution for detailed monitoring, while peripheral areas use coarser resolution. This localized adaptation ensures that computational resources are concentrated where they are most needed for safety-critical monitoring, rather than uniformly distributing cells across the entire grid.
3Ease of manufacture
If a square or rectangular grid is used to cover the region of interest, then the grid structure is simple, but many unnecessary cells are included that waste memory and slow down processing
Solution Approach 1:
The patent segments the occupancy grid into multiple zones based on the lane geometry and region of interest. The grid is divided into relevant areas (within lane boundaries and near the vehicle) and irrelevant areas (outside lane boundaries and far from the vehicle). By segmenting the grid in this manner, the system can selectively process only the necessary cells, eliminating unnecessary data while maintaining the fundamental grid structure for efficient computation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer implemented method is provided for generating a dynamic occupancy grid in front of a host vehicle. An indicator is detected for the course of a lane of a road in front of the host vehicle, and a base area is determined based on the detected indicator, wherein the base area is restricted to a region of interest in front of the host vehicle. A plurality of cells is defined by dividing the base area in order to form the occupancy grid. For each cell of the occupancy grid, it is determined whether the cell is occupied at least partly by a detectable object.