LiDAR Ground Plane Detection Using Projection Criteria
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional LiDAR data processing techniques, such as height thresholding, struggle to reliably differentiate between ground and objects/hazards due to dynamic vehicle motion and non-flat driving surfaces, leading to inaccurate environmental mapping for autonomous vehicles.
Innovation Solution
The proposed method processes LiDAR data by decomposing range measurements into height and ground projections, computing differentials, and applying a ground projection criterion and angular variation criterion to identify range measurements as ground or non-ground points, enhancing the discrimination between free space and obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If height thresholding is used to classify objects, then processing speed is maintained, but reliability of ground vs. object differentiation deteriorates due to dynamic vehicle motion and non-flat surfaces
Solution Approach 1:
The patent segments the LiDAR data processing by dividing the point cloud into ground points and non-ground points using two separate criteria: ground projection criterion (comparing projected positions with actual positions) and angular variation criterion (analyzing angular differences between adjacent points). This segmentation allows each criterion to focus on specific aspects of ground detection, improving overall reliability while keeping individual processing steps manageable
Solution Approach 2:
The patent introduces a new dimension of analysis by projecting 3D LiDAR points onto a 2D ground plane and then analyzing the angular variations in the projected data. This dimensional transformation enables the system to detect ground surfaces more reliably by examining spatial relationships in multiple dimensions rather than relying solely on height thresholding in a single dimension
2Measurement precision
If conventional height thresholding is applied, then processing simplicity is maintained, but measurement precision deteriorates in dynamic and non-flat environments
Solution Approach 1:
The patent replaces the simple mechanical height thresholding approach with a more sophisticated computational system that uses geometric projections and angular calculations. Instead of relying on a fixed height cutoff, the system substitutes mechanical simplicity with computational geometry, using ground projection to map points onto a reference plane and angular variation analysis to detect surface characteristics, thereby achieving higher measurement precision through mathematical substitution
3Reliability
If ground projection criterion and angular variation criterion are applied, then discrimination reliability improves, but computational complexity increases
Solution Approach 1:
The patent implements a two-stage processing approach where the ground projection criterion is applied first to identify potential ground points, and then the angular variation criterion is applied selectively to refine the classification. This partial application of criteria allows the system to achieve high discrimination reliability while reducing unnecessary computational effort on points that are clearly identified in the first stage
Solution Approach 2:
The patent changes the parameters used for ground detection from simple height values to composite parameters including projected position coordinates and angular variations. By transforming the input parameters and using multiple parameters simultaneously, the system achieves more reliable discrimination between free space and obstacles, with each parameter contributing specific information that complements the others
Data Source
AI summary
Improved processing of sensor data (e.g., LiDAR data) can be used to distinguish between free space and objects/hazards. Autonomous vehicles can use such information for performing autonomous driving and/or parking operations. LiDAR data can include a plurality of range measurements (e.g., forming a 3D point cloud). Each of the range measurements can correspond to a respective LiDAR channel and azimuth angle. The processing of LiDAR data can include identifying one or more of the plurality of range measurements as ground points or non-ground points based on one or more point criteria. The one or more point criteria can include a ground projection criterion and an angular variation criterion.