LiDAR Point Cloud Ground Filtering for Real-Time Multi-Scan Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LiDAR systems in autonomous vehicles face challenges with excessive data quantity and inconsistent filtering performance due to varying point cloud characteristics from multiple LiDARs, leading to inefficiencies in real-time data processing.
Innovation Solution
A point cloud processing device employing a processor that filters ground points using a random sample consensus algorithm for vertically scanned clouds and ray ground filtering for horizontally scanned clouds, followed by data fusion to generate a processed point cloud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple LiDARs are used to extract point clouds, then blind spots are reduced and safety is improved, but data quantity becomes excessively large
Solution Approach 1:
The patent extracts and removes ground points from the point cloud data through specialized filtering algorithms (RANSAC for vertical scans, ray ground filtering for horizontal scans). This separates the ground component from the point cloud, allowing retention of only the necessary non-ground data for autonomous driving decisions, thus reducing data quantity while maintaining safety.
Solution Approach 2:
The patent discards ground points that have been successfully filtered and removed from the point cloud. By identifying and discarding these redundant ground points through algorithmic processing, the system recovers computational resources and reduces the data burden while preserving critical non-ground information needed for safety-critical functions.
2Device complexity
If the same filtering algorithm is used for all point cloud data, then processing is simplified, but filtering performance deteriorates
Solution Approach 1:
The patent applies different filtering algorithms to different types of point cloud data based on their scanning characteristics. Vertical scans use RANSAC algorithm while horizontal scans use ray ground filtering. This local adaptation of filtering methods to match the specific characteristics of each scan type optimizes filtering performance for each case without requiring complex unified processing.
Solution Approach 2:
The patent changes the filtering approach based on the scanning parameters and characteristics of the LiDAR system. By detecting whether the scan is vertical or horizontal, the system dynamically selects the appropriate algorithm, adapting the processing method to the input data characteristics and achieving optimal filtering performance across different scenarios.
Data Source
AI summary
A point cloud processing device and a point cloud processing method are provided. The point cloud processing method includes: receiving a vertically scanned point cloud; performing a random sample consensus algorithm on the vertically scanned point cloud to obtain a fitting plane; filtering a first ground point of the vertically scanned point cloud according to the fitting plane to update the vertically scanned point cloud; and generating a processed point cloud according to the updated vertically scanned point cloud and outputting the processed point cloud.