LiDAR Dynamic Thresholding for Low-Reflectivity Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately detecting objects using LiDAR systems due to false detections caused by solar radiation, electrical noise, and varying environmental conditions, which are exacerbated by the difficulty in setting an appropriate detection threshold.
Innovation Solution
A dynamic detection threshold system for LiDAR sensors in autonomous vehicles, which includes a LiDAR system with a light emitter and detector, a comparator, and a controller. The controller adjusts the detection threshold based on aggregated digital output signals from the comparator, allowing for precise object detection and reduced false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a low detection threshold is used, then more objects including low-reflectivity objects can be detected, but false detections increase due to solar radiation, light from other light sources, and electrical noise
Solution Approach 1:
The patent implements a dynamic threshold adjustment mechanism where the detection threshold is not fixed but adapts based on environmental conditions. The system monitors ambient light levels, solar radiation, and electrical noise characteristics, then automatically adjusts the threshold to optimize detection accuracy while minimizing false positives. This resolves the contradiction by making the threshold responsive to changing conditions rather than static.
Solution Approach 2:
The system changes the detection threshold parameter dynamically based on measured environmental factors. By monitoring parameters such as ambient light intensity, solar radiation levels, and electrical noise characteristics, the system adjusts the threshold parameter to maintain optimal detection performance across varying conditions, resolving the trade-off between detecting low-reflectivity objects and avoiding false detections.
2Object-generated harmful factors
If a high detection threshold is used, then false detections are reduced, but certain objects such as low-reflectivity objects and distant objects are not detected
Solution Approach 1:
The dynamic threshold adjustment mechanism allows the system to use higher thresholds when environmental noise is high (reducing false detections) and lower thresholds when environmental noise is low (improving detection of difficult objects). This resolves the contradiction by adapting the threshold to current conditions rather than using a fixed high value.
Solution Approach 2:
The system incorporates feedback loops that continuously monitor detection results and environmental conditions, then adjust the threshold accordingly. When false detections are detected or environmental noise increases, the threshold is adjusted upward. When detection sensitivity needs improvement and conditions permit, the threshold is lowered, resolving the contradiction through continuous feedback-driven optimization.
3Ease of operation
If a fixed threshold is used, then the system is simple to operate, but it cannot adapt to varying environmental conditions throughout the day
Solution Approach 1:
The system performs self-adjustment of the detection threshold by automatically monitoring environmental conditions and modifying the threshold without user intervention. The controller autonomously adapts to varying solar radiation, ambient light, and electrical noise conditions throughout the day, maintaining optimal performance while keeping the operation simple for the user.
Solution Approach 2:
The threshold transitions from a static, manually-set value to a dynamic, automatically-adjusted parameter that adapts to environmental conditions. This dynamic behavior enables the system to respond to changing conditions throughout the day while maintaining ease of operation, as the adaptation occurs automatically without requiring user input or complex manual adjustments.
4Measurement precision
If digitization with high-speed ADCs is used, then precise amplitude estimation is achieved, but the system becomes very expensive and generates large amounts of useless data
Solution Approach 1:
The patent extracts only the essential information needed for detection (presence/absence of objects, basic range data) while discarding excessive data that would require high-speed ADCs. By using a dynamic threshold approach with analog-to-digital conversion only when necessary, the system achieves sufficient measurement precision without the cost and data processing burden of continuous high-speed digitization.
Solution Approach 2:
Instead of investing in expensive high-speed ADCs that generate large amounts of data, the system uses a more economical approach with lower-speed conversion triggered only when detection events occur. This disposable-like strategy of converting analog signals to digital only when necessary achieves the required measurement precision while dramatically reducing system cost and data processing requirements.
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
The dynamic threshold system enhances the accuracy of object detection by reducing false positives and improving the detection of low-reflectivity or distant objects, while also adapting to varying environmental conditions without the need for high-speed ADCs.
Implementation Method 1
at least one light detector configured to receive reflected pulses of light and generate analog output signals based on the reflected pulses of light
Implementation Method 2
A comparator may be configured to receive the analog output signals from the light detector and generate digital output signals based on the analog output signals and a threshold
Implementation Method 3
Certain LiDAR systems operate on a time-of-flight principle by measuring the time difference between emission of a signal and detection with a sensor of the signal returning after being reflected by an object
Data Source
AI summary
A system including a LiDAR system of an autonomous vehicle. The LiDAR system includes at least one light emitter and at least one light detector to generate analog output signals based on reflected pulses of light. A comparator receives the analog output signals from the light detector and generates digital output signals based on the analog output signals and a threshold. A controller receives a first digital output signal of the digital output signals from the comparator based on the threshold, adjusts the threshold, receives at least one further digital output signal of the digital output signals from the comparator based on the threshold as adjusted, and/or determines at least one aggregation based on the first digital output signal and the further digital output signal(s).