Autonomous Vehicle Obstacle Detection via Lidar-Map Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving systems face challenges in accurately detecting obstacles due to inconsistencies between lidar data and data from other sensors, leading to missed detections or false alarms, especially in diverse and dynamic environments.
Innovation Solution
A method that utilizes lidar point clouds in conjunction with 3D mapping and correlation analysis to resolve inconsistencies between sensor data, determining the presence of obstacles by comparing lidar measurements with expected road signatures and adjusting for errors, thereby improving detection robustness and reducing false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lidar data is used for obstacle detection, then detection precision is improved, but false detections increase due to inconsistencies with other sensor data
Solution Approach 1:
The patent combines lidar point cloud data with 3D mapping data and other sensor data (camera and radar) to perform joint obstacle detection. The fusion of multiple data sources allows the system to cross-validate detections, confirming true obstacles while filtering out false positives that appear in only one sensor modality, thus improving reliability while maintaining precision.
Solution Approach 2:
The system uses 3D mapping data as a reference framework to provide feedback on expected road infrastructure. By comparing actual lidar detections against the predetermined 3D map of road signs, barriers, and infrastructure, the system can identify inconsistencies and adjust detections accordingly, reducing false alarms while maintaining accurate obstacle detection.
2Reliability
If multiple sensor technologies are combined for obstacle detection, then detection robustness is improved, but device complexity increases
Solution Approach 1:
The patent makes the 3D mapping data serve multiple functions: it provides a reference framework for obstacle detection, validates lidar detections, reduces false positives, and compensates for limitations of individual sensors. This multi-functional use of existing mapping data improves detection robustness without adding new hardware complexity.
Solution Approach 2:
The system uses the vehicle's own 3D mapping data (already present in the vehicle's navigation system) to enhance obstacle detection capabilities. By leveraging existing onboard resources rather than requiring additional external systems, the patent improves detection robustness while avoiding increased device complexity.
3Measurement precision
If lidar point clouds are processed for obstacle detection, then measurement precision is improved, but computational cost increases
Solution Approach 1:
The patent pre-processes and stores 3D mapping data in advance, organizing it into a ready-to-use reference framework before the vehicle reaches the geographical area. This preliminary preparation allows the system to quickly compare incoming lidar point clouds against pre-organized map data during real-time detection, reducing computational energy requirements during actual obstacle detection operations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention concerns a robust method for detecting obstacles in a given geographical area. The method comprises a step of pre-detection by processing measurements taken in said area by several exteroceptive sensors including at least one lidar sensor and at least one non-lidar sensor, in particular of the radar or camera type. It further comprises a step of confirming the pre-detection step, including comparing a cloud of points measured by the lidar sensor in said area with map data corresponding to said area.