Microphone Array Filtering for Autonomous Vehicle Siren Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicle navigation technologies face challenges in accurately detecting objects using acoustic perception due to interference noises, leading to reduced accuracy and safety, especially in scenarios like emergency vehicle sirens where precise navigation is critical.
Innovation Solution
Implementing an unconventional microphone array and sound signal processing device that acts as a spatial sound signal frequency band pass filter, amplifying specific frequencies from the front or behind the vehicle while filtering out other directions, thereby reducing computational complexity and improving object detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If microphones are mounted on various locations of the autonomous vehicle to capture all sounds and noises, then the coverage of acoustic detection is improved, but the accuracy of object detection deteriorates due to interference noises
Solution Approach 1:
The patent segments the acoustic signal processing into multiple frequency bands using bandpass filters. Each frequency band is processed separately to identify specific sound sources (e.g., emergency vehicle sirens, horns, tire noises) while filtering out irrelevant frequencies. This segmentation allows the system to maintain broad detection coverage while achieving high accuracy for specific objects of interest.
Solution Approach 2:
The patent applies different processing qualities to different frequency ranges. Critical frequency bands containing important navigation information (emergency vehicle sounds, horn sounds) are amplified and processed with higher priority, while less relevant frequency bands are filtered or processed with lower computational resources. This local quality differentiation improves detection accuracy for critical objects without requiring uniform high-quality processing across all frequencies.
2Adaptability or versatility
If all captured sounds and noises are processed to detect objects on the road, then the comprehensiveness of object detection is improved, but the computational complexity increases
Solution Approach 1:
The patent divides the complex task of processing all captured sounds into separate frequency band processing tasks. Each bandpass filter handles a specific frequency range, and the system selectively processes only those bands containing relevant information for navigation (emergency vehicles, horns, tire noises). This segmentation reduces computational complexity by avoiding processing of irrelevant frequency ranges while maintaining comprehensive detection of important objects.
Solution Approach 2:
The patent extracts and processes only the specific frequency bands containing navigation-critical information (emergency vehicle sirens, horn sounds, tire noise patterns) while discarding or minimally processing other frequency ranges. This extraction approach maintains comprehensive detection of important objects without the computational burden of processing all captured sounds equally.
3Quantity of substance
If interference noises are included in the sound signal processing, then the completeness of acoustic data is improved, but the accuracy of identifying sound sources deteriorates
Solution Approach 1:
The patent segments the acoustic spectrum into multiple frequency bands and processes each band separately. This allows the system to maintain complete acoustic data across all frequencies while identifying sound sources with high accuracy in specific critical bands (emergency vehicle sirens, horns). The segmentation enables selective focus on frequency ranges containing important navigation information without losing overall acoustic completeness.
Solution Approach 2:
The patent applies different processing qualities to different frequency bands. Bands containing critical navigation information (emergency vehicle sounds, horn sounds) receive high-priority processing with amplification and detailed analysis, while other bands are processed with lower priority. This local quality differentiation maintains sound source identification accuracy for critical objects while preserving complete acoustic data coverage.
Data Source
AI summary
A system comprises a microphone array and a processor. The microphone array detects first sound signal and the second sound signal. Each of the first and second sound signals has a particular frequency band. Each of the first and second sound signals is originated from a particular sound source. The processor receives the first and second sound signals. The processor amplifies the first sound signals, each with a different amplification order. The processor disregards the second sound signal, where the second sound signal includes interference noise signals. The processor determines that the first sound signals indicate that a vehicle is within a threshold distance from an autonomous vehicle and traveling in a direction toward the autonomous vehicle. In response, the processor instructs the autonomous vehicle to perform a minimal risk condition operation. The minimal risk condition operation includes pulling over or stopping the autonomous vehicle.