Vehicle LiDAR Crosstalk Filtering Using Two-Pass Point Cloud Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LIDAR devices in motor vehicles suffer from crosstalk phenomena due to high-power light emissions, leading to erroneous object detection and disrupting vehicle navigation, particularly in autonomous vehicles, and existing solutions require expensive dedicated hardware.
Innovation Solution
A method using two-pass filtering to identify and exclude crosstalk-related points by measuring light intensity and distance thresholds, employing standard LIDAR hardware and minimal computational resources without additional components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Length of stationary object
If high-power light emissions are used to detect distant objects, then the detection range is improved, but crosstalk artifacts occur causing erroneous object detection
Solution Approach 1:
The method performs preliminary identification of crosstalk artifacts using distance and brightness thresholds before final object detection. By pre-identifying points that meet crosstalk criteria (bright nearby points with excessive brightness and distance relationships), the system removes these artifacts before they can cause erroneous detection, thus maintaining reliability while using high-power emissions for extended range
Solution Approach 2:
The invention changes the parameters used for object detection by introducing distance-based and brightness-based filtering criteria. By analyzing the distance between points and their relative brightness values, the system dynamically identifies and removes crosstalk artifacts. This parameter-based approach allows the use of high-power light emissions for extended detection range while maintaining detection accuracy through mathematical filtering
2Reliability
If conservative data filtering is applied to avoid false negatives, then safety is improved, but false positives increase leading to unexpected vehicle behavior
Solution Approach 1:
The method uses feedback mechanisms by continuously analyzing the spatial and brightness relationships between detected points. The system calculates distance thresholds and brightness ratios, using this feedback information to identify patterns characteristic of crosstalk artifacts. This feedback-driven approach allows the system to distinguish between real objects and artifacts dynamically, reducing false positives while maintaining safety through conservative filtering of identified artifacts
Solution Approach 2:
By performing preliminary identification and removal of crosstalk artifacts before final detection decisions, the system prevents false positives from being generated in the first place. This preliminary filtering action, based on distance and brightness criteria, allows the system to be less conservative in subsequent detection steps, thereby reducing false positives while maintaining safety
3Reliability
If dedicated hardware components are used to filter crosstalk artifacts, then detection accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The invention replaces potential hardware-based filtering solutions with a computational method. Instead of using dedicated hardware components to physically filter or block crosstalk signals, the system uses software-based algorithms that analyze point cloud data, calculate distance and brightness relationships, and identify artifacts through mathematical criteria. This substitution of mechanical/hardware filtering with computational filtering maintains detection accuracy while significantly reducing device complexity and cost
Solution Approach 2:
The LIDAR device uses its own existing data (point cloud, distance information, brightness values) to identify and remove its own artifacts without requiring external or dedicated filtering hardware. The system performs self-diagnosis and self-correction by analyzing its detection data and removing identified crosstalk artifacts, thereby maintaining accuracy without adding hardware complexity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Effectively filters crosstalk artifacts with negligible false negatives, ensuring reliable object detection and safe vehicle operation using conventional LIDAR devices.
Implementation Method 1
emitting an incident light pulse from the motor vehicle toward its external environment
Implementation Method 2
receiving in return, on a photodetector of the motor vehicle, a reflected light pulse
Implementation Method 3
receiving in return, on a photodetector of the motor vehicle, a reflected light pulse
Data Source
AI summary
A method for implementing a LIDAR device in a motor vehicle, including the following steps: determining the number of bright nearby points of the point cloud, a bright nearby point being defined as a point that meets the following two conditions: the distance of separation corresponding to the point is less than a first predetermined distance threshold; and the light intensity corresponding to the point is higher than a first predetermined light intensity threshold; when the number of bright nearby points is higher than a predetermined crosstalk filtering activation threshold, identifying each crosstalk-related point that meets the following two conditions: the distance of separation corresponding to the point is less than a second predetermined distance threshold; and the light intensity corresponding to the point is lower than a second predetermined brightness threshold.