Ground Intensity LiDAR Localization for Tunnel Pose Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional localization systems for autonomous vehicles struggle in areas with insufficient geometric constraints, such as tunnels and bridges, where geometric-based methods fail to provide accurate pose estimates due to lack of diverse geometric features, leading to multiple valid solutions.
Innovation Solution
A ground intensity LIDAR localizer is introduced that utilizes LIDAR intensity information to complement geometric localization, enabling simultaneous operation of multiple localizers to provide accurate 6-degrees-of-freedom pose estimation by aligning ground intensity LIDAR images using registration algorithms like Gauss-Newton or Levenberg-Marquardt frameworks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional geometric-based localization systems are used, then localization works well in environments with diverse geometric features, but localization accuracy deteriorates in geometrically degenerate areas such as tunnels, bridges, and highways
Solution Approach 1:
The patent combines geometric LIDAR data with intensity LIDAR data to create a hybrid localization system. The geometric localizer processes 3D point cloud geometry while the intensity localizer processes reflectance intensity values, and both localizers are fused to provide comprehensive pose estimation that works across diverse environments including geometrically degenerate areas.
Solution Approach 2:
The patent adds intensity as an additional dimension to the traditional geometric LIDAR data. Instead of relying solely on 3D spatial coordinates (x, y, z), the system incorporates intensity values (reflectance properties) as a fourth dimension, enabling localization in environments where geometric features are insufficient.
2Reliability
If multiple localizers operate simultaneously, then localization robustness improves across different environments, but system complexity increases
Solution Approach 1:
The patent divides the localization system into separate functional modules: a geometric localizer that processes spatial geometry and an intensity localizer that processes reflectance values. Each localizer is independently optimized for its specific data type, and their results are fused at the pose estimation level, reducing interdependencies and simplifying system management.
Solution Approach 2:
The patent creates a unified localization framework that can handle both geometric and intensity data through a common pose estimation and fusion architecture. The system uses a single map-building process that incorporates both data types and a unified localization pipeline that processes both localizers, reducing overall system complexity despite the multiple localizers.
Data Source
AI summary
A system for determining a pose of a vehicle and building maps from vehicle priors processes received ground intensity LIDAR data including intensity data for points believed to be on the ground and height information to form ground intensity LIDAR (GIL) images including pixels in 2D coordinates where each pixel contains an intensity value, a height value, and x- and y-gradients of intensity and height. The GIL images are formed by filtering aggregated ground intensity LIDAR data falling into a same spatial bin on the ground and using a registration algorithm to align two GIL images relative to one another by estimating a 6-degree-of-freedom pose with associated uncertainty that minimizes error between the two GIL images. The aligned GIL images are provided as a pose estimate to a localizer. The system may provide online localization and pose estimation, prior building, and prior to prior alignment pose estimation using image-based techniques.