FMCW LiDAR Point Cloud Filtering for Ghost Point Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
FMCW LiDAR systems suffer from phase impairments leading to ghost points or noisy points that introduce errors in target range and velocity estimation due to incorrect peak matching, causing false alarms.
Innovation Solution
A point cloud filtering method that identifies and removes false alarm points by exploiting their distinct characteristics using neighborhood context and metrics, allowing for the selection of neighborhood points to filter out ghost points while preserving true detections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If FMCW LiDAR systems perform peak matching for target detection, then target detection capability is improved, but phase impairments cause ghost points and false alarms that reduce measurement precision
Solution Approach 1:
The patent segments the point cloud data into multiple regions or groups based on spatial distribution and characteristics. By dividing the point cloud into smaller subsets, the system can apply filtering operations more effectively to identify and remove ghost points while preserving true target detections, thus resolving the contradiction between detection capability and measurement precision.
Solution Approach 2:
The patent introduces intermediary filtering operations between peak matching and final target detection. These filtering steps act as mediators that process the raw point cloud data, removing false alarm points and ghost points before the final target identification, thereby improving measurement precision without compromising detection capability.
2Measurement precision
If filtering operations are applied to remove ghost points, then measurement precision is improved, but system complexity increases due to additional processing steps
Solution Approach 1:
The patent applies filtering operations with local quality by adapting the filtering strength and parameters to different regions of the point cloud. Instead of applying a uniform filtering approach, the system adjusts filtering characteristics based on local point density, spatial distribution, and detected anomaly patterns, improving precision while managing complexity through localized rather than global processing.
Solution Approach 2:
The patent utilizes parameter changes in the filtering algorithm to balance precision and complexity. By dynamically adjusting filtering parameters such as threshold values, neighborhood sizes, and confidence levels based on the input data characteristics, the system achieves high measurement precision while avoiding excessive complexity through adaptive parameter selection rather than fixed complex operations.
3Reliability
If multiple filters are used to identify different types of false alarm points, then false alarm reduction is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by implementing a hierarchical filtering approach where simpler, faster filters are applied first to remove obvious false alarms and ghost points. Subsequent more complex filters are then applied only to the remaining points that require more sophisticated analysis, thereby reducing overall processing time while maintaining high false alarm reduction effectiveness.
Solution Approach 2:
The patent employs partial action by selectively applying different filtering operations to different portions of the point cloud based on their characteristics. Instead of applying all filters uniformly to all points, the system identifies regions with high false alarm probability and applies intensive filtering only to those areas, reducing total processing time while maintaining high reliability in critical regions.
Data Source
AI summary
A set of POIs of a point cloud are received at a first filter, where each POI of the set of POIs comprises one or more points. Each POI of the set of POIs is filtered. A set of neighborhood points of a POI is selected. A metric for the set of neighborhood points is computed based on a property of the set of neighborhood points and the POI, wherein the property comprises a velocity. Based on the metric, whether to accept the POI, modify the POI, reject the POI, or transmit the POI to a second filter, to extract at least one of range or velocity information related to the target is determined. Provided the POI is not accepted, modified, or rejected, the POI is transmitted to the second filter to determine whether to accept, modify, or reject the POI to extract the at least one of range or velocity information related to the target.