A Noise-Robust Method with Smoothedℓ1/ℓ2Regularization for Sparse Moving-Source Mapping
| Autor: | Jrme I. Mars, Mai Quyen Pham, Benoit Oudompheng, Barbara Nicolas |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Blind deconvolution
Beamforming Computer science Spectral density 020206 networking & telecommunications 02 engineering and technology Sparse approximation 01 natural sciences Regularization (mathematics) Noise Control and Systems Engineering Frequency domain 0103 physical sciences Signal Processing 0202 electrical engineering electronic engineering information engineering Computer Vision and Pattern Recognition Time domain Deconvolution Electrical and Electronic Engineering 010301 acoustics Algorithm Software |
| Zdroj: | Signal Processing. 135:96-106 |
| ISSN: | 0165-1684 |
| DOI: | 10.1016/j.sigpro.2016.12.022 |
| Popis: | The method described here performs blind deconvolution of the beamforming output in the frequency domain. To provide accurate blind deconvolution, sparsity priors are introduced with a smoothed 1/2 regularization term. As the mean of the noise in the power spectrum domain depends on its variance in the time domain, the proposed method includes a variance estimation step, which allows more robust blind deconvolution. Validation of the method on both simulated and real data, and of its performance, are compared with two well-known methods from the literature: the deconvolution approach for the mapping of acoustic sources, and sound density modeling. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect