LiDAR Echo Denoising with Position- and Distance-Based Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LiDAR denoising methods use a single threshold to remove noise signals, which limits denoising accuracy and results in suboptimal performance due to varying noise types and echo signal energies across different regions of the receiving array.
Innovation Solution
A LiDAR denoising method that acquires echo signals for pixel units, obtains target measurement distance and echo feature information, and adjusts noise point thresholds based on pixel position and distance to accurately identify and remove noise signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single threshold is used to remove noise signals, then the denoising process is simple, but the denoising accuracy is limited and performance is suboptimal
Solution Approach 1:
The patent applies local quality by dividing the receiving array into multiple regions and assigning different noise point thresholds to each region. Each region's threshold is optimized for its specific noise characteristics, allowing the system to achieve high denoising accuracy across different spatial locations without requiring overly complex global processing.
Solution Approach 2:
The patent changes the threshold parameter dynamically based on the pixel position in the receiving array. By adjusting the noise point threshold according to the specific region and target measurement distance, the system adapts to varying noise levels and echo signal energies across different areas, resolving the contradiction between simple processing and high accuracy.
2Measurement precision
If different noise point thresholds are used for different pixel positions and distances, then denoising accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent segments the receiving array into multiple regions and pre-determines noise point thresholds for each region. This segmentation approach allows the system to manage complexity by organizing thresholds in a structured manner, where each region has its own optimized threshold based on characteristic noise levels and echo signal energies, making the complex threshold adjustment manageable and systematic.
Solution Approach 2:
The patent performs preliminary determination of noise point thresholds for different regions before actual denoising operations. By pre-calculating and storing region-specific thresholds based on characteristic noise levels and echo signal energies, the system avoids complex real-time calculations during operation, thus improving denoising accuracy while keeping operational complexity low.
Data Source
AI summary
A LiDAR denoising method and a LiDAR are provided. The LiDAR denoising method includes: first obtaining the echo signal corresponding to a pixel unit in a receiving array of the LiDAR; obtaining a target measurement distance and echo feature information based on the echo signal, where the echo feature information is an echo amplitude and/or an echo width; obtaining a target noise point threshold based on a pixel position of the pixel unit and the target measurement distance; determining whether the echo signal is a noise signal according to the echo feature information and the target noise point threshold; and if so, deleting the echo signal.