11:10
Drone Audition - Listening with Drones
11:10
20 mins
|
UAV sound monitoring system for bird survey
Lin Wang
Abstract: Unmanned aerial vehicles (UAVs) are increasingly utilized in ecological and environmental research to study sensitive wildlife in remote or inaccessible areas. This is particularly relevant for monitoring bird species that inhabit forest canopies or regions with limited human access. UAV-based bioacoustic monitoring offers a promising alternative to traditional ground-based methods for surveying songbird populations. This paper presents the development of a low-cost UAV platform for in-flight acquisition of bird call data. The performance of several microphone types is analyzed with respect to their ability to reduce UAV-generated noise during flight. Field experiments conducted in real-world environments demonstrate the successful recording of bird vocalizations at elevated altitudes. The results validate the feasibility of the proposed system and indicate its potential for future bird species detection and bioacoustic analysis.
|
11:30
20 mins
|
Development of a Hybrid Denoising Methodology for UAV Onboard Acoustic Monitoring: Coupling Passive Shielding and Adaptive Filtering
Florent Masson, Sylvain Levent, Eliot Naviere, Xavier Watremez
Abstract: The integration of acoustic sensors on Unmanned Aerial Vehicles (UAVs) offers significant potential for industrial condition monitoring. However, operational deployment is severely restricted by the high-intensity "ego-noise" generated by the drone's propulsion system, which masks environmental acoustic signatures. This paper presents the development and validation of a Proof of Concept (PoC) designed to isolate target sounds from drone inherent noise using a hybrid methodology combining hardware design and adaptive signal processing.
The methodology relies on a specific acquisition prototype embedded on a Holybro X500 V2 platform. The hardware architecture uses a dual-microphone setup: a reference microphone captures the drone’s noise, while an observation microphone records the target scene. To enhance the Signal-to-Noise Ratio (SNR) physically, a custom 3D-printed casing featuring specific cones profiles was developed to provide passive acoustic shielding and directivity control.
Regarding signal processing, two strategies were evaluated: Wiener filtering and the Normalized Least Mean Squares (NLMS) algorithm. Results demonstrate that while Wiener filtering offers superior suppression for stationary signals, the NLMS algorithm is required for dynamic flight conditions. The study identified that a minimum physical separation and passive filtering are critical enablers for the algorithm's stability. The combined approach achieved an average noise reduction of more than 15 dB in the 100–4000 Hz range. While challenges remain with high-frequency harmonics and complex signal preservation, this research validates the hybrid mechanical-digital approach as a prerequisite for onboard acoustic listening.
|
11:50
20 mins
|
Listening with drones: Denoising algorithms for best performance
Paula Chale Castell, Roberto Merino-Martinez, Toros Senan, Stephy Annie Curie, Mirjam Snellen
Abstract: Unmanned Aerial Vehicles (UAVs) have emerged as indispensable assets in modern search-and-rescue operations, providing rapid situational awareness in environments where ground access is restricted. While optical and thermal sensors are standard, their effectiveness is often compromised by occlusion, dense vegetation, or adverse visual conditions, such as smoke and darkness. Drone audition, i.e. the use of embedded microphone arrays to detect and localize victims, offers a promising complementary modality. However, the practical deployment of this technology is currently hindered by severe acoustic interference from the UAV’s propulsion system, typically known as ego-noise, which drastically degrades the Signal-to-Noise Ratio (SNR) and masks target signals.
This research presents a comprehensive framework for evaluating and enhancing acoustic detection from aerial platforms. The methodology adopts a rigorous, two-stage validation approach designed to bridge the gap between theoretical algorithms and real-world applicability. In the first phase, we conduct a comparative benchmark of diverse noise suppression and Direction-of-Arrival (DoA) estimation architectures using previously recorded acoustic datasets. This establishes a performance baseline against ground-truth data, allowing for the optimization of filter parameters prior to hardware implementation.
Subsequently, the study transitions to experimental validation using a microphone array and a DJI Mavic 3 Enterprise drone. These experiments are conducted within the anechoic chamber of Delft University of Technology to simulate ideal free-field sound propagation. This controlled environment effectively isolates the system from environmental reverberations and multipath effects, ensuring that acoustic interference is exclusively attributable to the drone’s ego-noise profile. By systematically analyzing the trade-offs between localization accuracy, noise rejection capabilities, and computational cost across different processing strategies, this study aims to identify robust acoustic sensing configurations suitable for integration into autonomous rescue systems.
|
12:10
20 mins
|
On Cascaded Ego-Noise Suppression for Drone-Based Sound Pressure Level Measurement
Steffen Büchholz, Ennes Sarradj
Abstract: Measuring sound pressure level at locations that are difficult to access from the ground is challenging. Drone-mounted microphones offer promising capabilities to address these challenges for locations such as elevated noise source positions or on envelope surfaces around industrial equipment. However, the ego-noise generated by the drone's rotors presents a fundamental problem, as it contaminates the acoustic signals of interest. This work addresses ego-noise suppression by developing a cascaded signal processing approach that separately targets the distinct tonal and broadband components of multicopter self-noise.
The proposed methodology exploits the spectral characteristics of rotor noise: strong tonal components at blade passing frequencies and their harmonics, combined with lower-power broadband contributions from turbulent flow interactions. For tonal suppression, adaptive IIR notch filters track the time-varying blade passing frequencies of each rotor. Filter center frequencies are updated either through direct rotational speed measurements or acoustic estimation from the microphone signals. Subsequently, spatial filtering techniques based on microphone array processing separate remaining broadband ego-noise from the measurement signal, exploiting the known directions of arrival from the rotor positions.
Experimental validation is performed in a controlled anechoic chamber environment using microphone array measurements of multicopter noise. We evaluate algorithm performance by analysing processed signals from scenarios with known acoustic conditions. Initial results demonstrate the cascaded approach's effectiveness in recovering the power spectrum across varying signal-to-ego-noise ratios. We will discuss trade-offs between notch filter adaptation strategies and reconstruction of spectral components removed during tonal filtering.
|
|