Dans cette thèse, on s'intéresse au développement de nouvelles méthodes d'auscultation GPR pour déterminer la géométrie et la structure des chaussées. Cette thèse a deux objectifs principaux. Tout d'abord, elle a pour but d'améliorer la compréhension des mécanismes de diffusion à très large bande dans un milieu stratifié composé d'interfaces rugueuses. Avec l'augmentation des fréquences d'utilisation de différents systèmes, les interfaces de chaussée ne peuvent plus être considérées comme planes. Ainsi, la rugosité des interfaces doit être prise en compte dans la modélisation de la propagation. Donc, une analyse de l'influence de cette rugosité sur l'onde rétrodiffusée a été réalisée. Elle a permis de montrer que la rugosité induit une décroissance en fréquence de l'amplitude des échos. Cette décroissance a ensuite été introduite dans le modèle du signal. Dans un second temps, plusieurs méthodes de traitement de signal ont été proposées pour estimer conjointement les paramètres de rugosité et d'épaisseur. D'abord, des méthodes multidimensionnelles ont été proposées en prenant en compte l'influence de la rugosité. Ensuite, afin de réduire la charge de calcul, des méthodes monodimensionnelles ont été proposées. Ces méthodes ont été évaluées à partir de signaux simulés. Les résultats ont montré de bonnes performances pour l'estimation des temps de retard et des paramètres de rugosité des interfaces. Enfin, les méthodes de traitement proposées dans ce manuscrit ont été testées sur des données expérimentales, qui permettent de valider les résultats théoriques et de montrer la faisabilité de la mesure de couches minces de chaussée et du paramètre de rugosité. / In this thesis, we focus on the development of new GPR methods to estimate the pavement structure. This thesis has two main objectives. First, it aims to improve the understanding of the scattering mechanisms for large-band radars in a stratified medium composed of rough interfaces. With increasing frequencies, pavement interfaces can no longer be considered as flat. The interface roughness must be taken into account in the propagation modelling. Thus, the influence of the roughness has been analysed. It has been shown that the interface roughness provides a continuous frequency decay of the magnitude of the echoes. This continuous frequency decay has then been introduced into the signal model. Secondly, several signal processing methods have been proposed to jointly estimate the roughness and thickness of pavement. Thus, multidimensional methods have been proposed by taking into account the roughness.Then, in order to reduce the computational burden, one-dimensional methods have also been proposed. From simulations, it can be seen that the proposed algorithms provide a good performance in parameter estimations (time delay, permittivity, roughness and thickness). Finally, the proposed signal processing methods are tested on experimental data. The results confirm the theoretical prediction. They show the feasibility to estimate both the thickness of thin pavements and roughness parameter.

In speech communication systems such as voice-controlled systems, hands-free mobile telephones and hearing aids, the received signals are degraded by room reverberation and background noise. This degradation can reduce the perceived quality and intelligibility of the speech, and decrease the performance of speech enhancement and source localization. These problems are difficult to solve due to the colored and nonstationary nature of the speech signals, and features of the Room Impulse Response (RIR) such as its long duration and non-minimum phase. In this dissertation, we focus on two topics of speech enhancement and speaker localization in noisy reverberant environments. A two-stage speech enhancement method is presented to suppress both early and late reverberation in noisy speech using only one microphone. It is shown that this method works well even in highly reverberant rooms. Experiments under different acoustic conditions confirm that the proposed blind method is superior in terms of reducing early and late reverberation effects and noise compared to other well known single-microphone techniques in the literature. Time Difference Of Arrival (TDOA)-based methods usually provide the most accurate source localization in adverse conditions. The key issue for these methods is to accurately estimate the TDOA using the smallest number of microphones. Two robust Time Delay Estimation (TDE) methods are proposed which use the information from only two microphones. One method is based on adaptive inverse filtering which provides superior performance even in highly reverberant and moderately noisy conditions. It also has negligible failure estimation which makes it a reliable method in realistic environments. This method has high computational complexity due to the estimation in the first stage for the first microphone. As a result, it can not be applied in time-varying environments and real-time applications. Our second method improves this problem by introducing two effective preprocessing stages for the conventional Cross Correlation (CC)-based methods. The results obtained in different noisy reverberant conditions including a real and time-varying environment demonstrate that the proposed methods are superior compared to the conventional TDE methods. / Graduate / 0544 / 0984 / saeed.mosayyebpour@gmail.com

In speech communication systems such as voice-controlled systems, hands-free mobile telephones and hearing aids, the received signals are degraded by room reverberation and background noise. This degradation can reduce the perceived quality and intelligibility of the speech, and decrease the performance of speech enhancement and source localization. These problems are difficult to solve due to the colored and nonstationary nature of the speech signals, and features of the Room Impulse Response (RIR) such as its long duration and non-minimum phase. In this dissertation, we focus on two topics of speech enhancement and speaker localization in noisy reverberant environments. A two-stage speech enhancement method is presented to suppress both early and late reverberation in noisy speech using only one microphone. It is shown that this method works well even in highly reverberant rooms. Experiments under different acoustic conditions confirm that the proposed blind method is superior in terms of reducing early and late reverberation effects and noise compared to other well known single-microphone techniques in the literature. Time Difference Of Arrival (TDOA)-based methods usually provide the most accurate source localization in adverse conditions. The key issue for these methods is to accurately estimate the TDOA using the smallest number of microphones. Two robust Time Delay Estimation (TDE) methods are proposed which use the information from only two microphones. One method is based on adaptive inverse filtering which provides superior performance even in highly reverberant and moderately noisy conditions. It also has negligible failure estimation which makes it a reliable method in realistic environments. This method has high computational complexity due to the estimation in the first stage for the first microphone. As a result, it can not be applied in time-varying environments and real-time applications. Our second method improves this problem by introducing two effective preprocessing stages for the conventional Cross Correlation (CC)-based methods. The results obtained in different noisy reverberant conditions including a real and time-varying environment demonstrate that the proposed methods are superior compared to the conventional TDE methods. / Graduate / 2015-04-23 / 0544 / 0984 / saeed.mosayyebpour@gmail.com