Parallel Cascade Point Cloud Filtering for Ghost Point Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
FMCW LiDAR systems suffer from ghost points or noisy points, which are falsely detected and introduce errors in target range/velocity estimation due to incorrect peak matching, leading to ghost objects in the scene.
Innovation Solution
Implement parallel cascaded filters that exploit characteristic features of false alarm points to identify and remove them by using multiple filters in parallel, each scoring a point of interest based on different metrics, and merging scores to make a decision on acceptance, modification, or rejection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple filters are used in parallel to identify false alarm points, then the accuracy of filtering is improved, but the device complexity increases
Solution Approach 1:
The filtering system is divided into multiple independent filters that operate in parallel. Each filter focuses on specific characteristic features of false alarm points (such as intensity, range, velocity, or spatial distribution), allowing the complex filtering task to be segmented into manageable, specialized components that collectively achieve high accuracy
Solution Approach 2:
The parallel filter structure creates a universal filtering system that can handle multiple types of false alarm patterns simultaneously. Each filter serves a specific function but collectively they provide multi-functional capability to address various noise patterns in point cloud data from different sources and conditions
2Reliability
If multiple filters process points independently to increase confidence, then the reliability of detection is improved, but the processing time increases
Solution Approach 1:
The parallel filter architecture enables continuous processing of point cloud data through multiple filters simultaneously. Instead of sequential processing where each filter waits for the previous one, all filters operate continuously and independently on the same input data, maintaining high throughput while achieving reliable detection through multiple independent assessments
Solution Approach 2:
The system applies multiple filtering operations (excessive action) to the same data points in parallel. This redundancy ensures that false alarm points are detected with high confidence by multiple filters, and the results can be merged efficiently to produce a reliable final classification without requiring all filters to complete sequentially
Data Source
AI summary
A set of POIs of a point cloud are received at a first filter. Each POI of the set of POIs is filtered. At a second filter, based on a first metric, a first score of the POI is determined. At a third filter, based on a second metric, a second score of the POI is determined. At the first filter, based on the first score and the second score, whether to accept the POI, modify the POI, or reject the POI, is determined to extract range or velocity information.