Gabriel Pablo Nava

Image processing techniques for high speed camera-based free-field optical communication

Optical communication through light-emitting diodes (LEDs) and video cameras is rapidly gaining attention due to the increasing pervassiveness of those devices, and because of its potential data capacity. However, the communication quality is greatly compromised by the low resolution of the imaging sensor which produces blurred images of the LEDs at long distances. On the other hand, the images recorded at high frame rates possess particular features that can be used to improve the reception. This paper suggests image processing techniques to detect and decode the optical signals of an array of LEDs. For the case of blurred images, detection of the LEDs is reinforced by a k-means clustering algorithm based on distance measurements derived from the linear correlation among pixel intensities along the time dimension. Experiments with a prototype show that the proposed algorithms can improve the bit error rate (BER) of the decoded signal. Furthermore, partial implementation on a General Purpose Graphics Processing Unit (GPGPU) is also addressed, and processing times are demonstrated.

A High-Speed Camera-Based Approach to Massive Sound Sensing With Optical Wireless Acoustic Sensors

This paper introduces an optical wireless audio system, which takes advantage of the parallel transmission feature offered by the arrangement of LEDs and a high-speed video camera. The motivation for building such system is the limitation encountered when deploying huge arrays of sound sensors, as existent wired and RF wireless microphones are undermined by complexity, energy, and bandwidth issues. In contrast, the proposed prototype is able to simultaneously capture 120 audio channels with full bandwidth each. Although the scalability to more channels is currently constrained by the camera interface hardware, our numerical analysis suggests that the proposed algorithms can decode optical signals of over 12 500 audio channels in a single GPU card. We discuss the various challenges involved in its deployment from the optical, image, and audio signal processing standpoints, and develop theoretical and practical solutions to accomplish a full real-time system integration. We further present examples of typical applications such as acoustic imaging of sound fields by beamforming with microphone arrays.

A new loudspeaker design for the enhancement of sound image localization on flat display panels

In most audio‐visual multimedia applications, conventional stereo loudspeakers have been used to implement auditory displays. However, a fundamental problem with this kind of displays is that only the listeners situated at the sweet spot and over the symmetrical axis of the loudspeaker array are able to accurately localize the sound images. Although a number of audio signal processing algorithms have been proposed to expand the listening area, relatively less study on new loudspeaker configurations has been explored. This paper introduces a simple, yet effective, loudspeaker design to enhance the localization of sound images over the surface of flat display panels. In contrast to previous approaches, expansion of the listening space is achieved by attachment of rigid barriers which physically modify the sound radiation pattern of the loudspeakers. Moreover, numerical simulations, experimental sound measurements, and subjective tests have been preformed to validate a prototype of the proposed loudspeaker d...

On the in situ measurement of acoustic impedance of the surfaces of a real room by an inverse optimization acoustical approach.

As numerical simulations of large‐scale complex‐shape sound fields have become affordable by powerful computers and algorithms, there is an increasing interest on new methods of in situ assessment of acoustic impedance and absorption coefficients to yield accurate sound predictions. Among the existent approaches, those based on inverse boundary frameworks offer the advantage of dealing with surfaces of arbitrary shapes. Moreover, estimation of the acoustic impedance of not one but all the surfaces in an interior using one microphone has been demonstrated. A drawback of these frameworks is a high sensitivity to noise due to a rank deficient inverse model. The authors have previously introduced an iterative optimization method that avoids the direct solution of such ill‐conditioned model by exploiting prior knowledge of the surface geometry. Numerical simulations suggest that the proposed approach improves the signal‐to‐noise ratio, which motivated preliminary experiments in a reverberation chamber. This re...

A Loudspeaker-Embedded Design for an Immersive Videoconferencing System with Large Screens

Traditional videoconferencing systems that work on the basis of face-to-face configurations allow the use of standard 2-channel stereo loudspeakers in order to achieve sound-image localization by the users. In contrast, immersive systems that do not constraint the users to a face-to-face viewpoint introduce new challenges in their acoustic design. This paper discusses the major problems that appear in a novel videoconferencing system in which the use of ordinary 2-channel loudspeakers is not suitable to achieve sound-image localization due to the inherent characteristics of the room space. In efforts to overcome these problems, a loudspeaker-embedded design is proposed and demonstrated with numerical simulations and experiments that show the steady-state sound field radiated by this new loudspeaker system.

In situ estimation of acoustic impedance on the surfaces of realistic interiors: An inverse approach

In situ measurement of acoustic impedance is traditionally performed using pairs of microphones located close to the test surface. However, this method becomes troublesome if inaccessible complex‐shaped surfaces, such as those in a real room, are considered. To overcome this problem, a method to estimate the normal acoustic impedance on the interior surfaces of a room is proposed. As input data, the algorithm takes: (1) The 3‐D shape of the room; (2) the strength of the sound source; and (3) a set of sound pressures measured at random locations in the interior sound field. The estimation of the acoustic impedance at each surface is achieved via the solution of an inverse problem, which arises from the boundary‐element method applied to the discretized interior boundaries of the room. Unfortunately, the solutions of this kind of problems are known to be unstable and sensitive to noise, due to a rank‐deficient linear system. Dealing with such a system is avoided in the proposed method by formulating an iter...