LiDAR Noise Point Identification Using Reflectivity and Continuity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LiDAR systems face challenges in identifying and filtering out noise points, particularly due to crosstalk between radars and adverse weather conditions like rain, snow, or fog.
Innovation Solution
A method for identifying noise points in LiDAR systems involves receiving a point cloud, obtaining reflectivity and continuity parameters for each point, and determining whether the point is a noise point based on predefined thresholds and confidence levels calculated from these parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If dual-pulse laser encoding is used to distinguish between multiple radars, then the ability to differentiate radar signals is improved, but crosstalk between radars still occurs when encoding intervals coincide
Solution Approach 1:
The patent changes the encoding parameter from fixed time intervals to variable time intervals. Each radar generates encoding intervals dynamically within a range (e.g., 1-5 nanoseconds), making it highly unlikely that two radars will generate the same encoding interval. This parameter variation effectively prevents crosstalk while maintaining signal differentiation capability.
2Measurement precision
If multiple laser beams are emitted in the vertical field of view to increase angular resolution, then the density of laser point cloud is improved, but the complexity of preventing crosstalk between channels increases
Solution Approach 1:
The patent merges the channel identification function into the time encoding mechanism. Each channel's laser pulses are assigned unique time encoding intervals, so the encoding process simultaneously achieves both signal differentiation and channel identification. This merging eliminates the need for separate channel tracking systems, reducing overall system complexity despite having multiple vertical beams.
3Object-generated harmful factors
If random encoding intervals are used between dual-pulse lasers, then the probability of crosstalk is reduced, but the reliability of decoding remains insufficient when encoding intervals are too similar
Solution Approach 1:
The patent implements a feedback mechanism where the receiving end detects the actual time interval between dual-pulse lasers and uses this detected interval for decoding. The system continuously adjusts and optimizes the decoding process based on the actual received encoding intervals, ensuring high decoding reliability even when radars operate with similar time ranges. This feedback loop maintains reliability while allowing flexible encoding interval selection.
Data Source
AI summary
A method for processing a point cloud generated by a light detection and ranging system, includes: receiving a first measurement generated by the light detection and ranging system; retrieving a plurality of second measurements that are adjacent to the first measurement; calculating a first parameter by using distances of the first measurement and the plurality of the second measurements, the first parameter indicating a degree of continuum of the first measurement relative to the plurality of the second measurements; and determining whether the first measurement represents a measurement of noises by using the first parameter.