LiDAR Velocity Grids for Curved-Trajectory Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating velocity grids for autonomous vehicles are not precise, leading to inaccurate object movement tracking, especially in complex scenarios like curved trajectories.
Innovation Solution
A simulated environment is used to create velocity grids by collecting and transforming LiDAR data from simulated objects, calculating their velocity, and storing this data in a grid, which improves the accuracy of object tracking and training for autonomous vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual methods are used to generate velocity grids, then the process is simple and quick, but the precision and accuracy of velocity estimates are poor
Solution Approach 1:
The patent uses simulated LiDAR data as a copy of real LiDAR data to generate velocity grids. By creating virtual representations of road scenarios with simulated objects and their corresponding LiDAR point clouds, the system can generate precise velocity estimates without the complexity of manual annotation. The simulated velocity grids are then used to train machine learning models that can process real data.
Solution Approach 2:
The patent performs preliminary velocity grid generation in a simulated environment before applying the methodology to real data. By pre-calculating velocity grids using simulated LiDAR data and known ground truth velocities, the system prepares training datasets that capture complex motion patterns without requiring manual annotation of real-world data.
2Reliability
If manual methods are used to generate velocity grids, then the setup is straightforward, but the accuracy in complex scenarios like curved trajectories is insufficient
Solution Approach 1:
The patent creates simulated copies of complex road scenarios including curved trajectories, intersections, and various object movements. These simulated scenarios replicate real-world complexity while maintaining known ground truth velocities, enabling accurate velocity grid generation for training purposes without the measurement errors inherent in manual methods.
3Measurement precision
If simulated environments are used to generate velocity grids, then the precision and accuracy of velocity estimates improve, but the complexity of the system increases
Solution Approach 1:
The patent uses simulated LiDAR data as a proxy for real data, creating virtual training datasets that are easier and safer to generate. The simulated velocity grids serve as ground truth labels for training machine learning models, eliminating the need for complex manual annotation processes while maintaining high precision.
Solution Approach 2:
The patent replaces manual mechanical annotation processes with automated simulation-based velocity grid generation. Instead of manually tracking objects and calculating velocities, the system uses simulated environments where velocities are known from the simulation physics, automatically generating precise velocity grids for training datasets.
Data Source
AI summary
A computed-implemented method is provided for creating a velocity grid using a simulated environment for use by an autonomous vehicle. The method may include simulating a road scenario with simulated objects, collecting first LiDAR data from simulated LiDAR sensors in the simulated road scenario, wherein the collected first LiDAR data comprises a first plurality of points that are representative of a first simulated object at a first 3D location and a first time. The method may also include transforming the first plurality of points from a simulated-scene frame-of-reference to a first simulated object frame-of-reference, and simulating the first simulated object to move from the first 3D location to a second 3D location within the simulated road scenario between the first time and a second time.