We introduce a novel long-range traffic monitoring system for vehicle detection, tracking, and classification based on fiber-optic distributed acoustic sensing (DAS). High resolution and long range are provided by the use of an optimized setup incorporating pulse compression, which, to our knowledge, is the first time that is applied to a traffic-monitoring DAS system. The raw data acquired with this sensor feeds an automatic vehicle detection and tracking algorithm based on a novel transformed domain that can be regarded as an evolution of the Hough Transform operating with non-binary valued signals. The detection of vehicles is performed by calculating the local maxima in the transformed domain for a given time-distance processing block of the detected signal. Then, an automatic tracking algorithm, which relies on a moving window paradigm, identifies the trajectory of the vehicle. Hence, the output of the tracking stage is a set of trajectories, each of which can be regarded as a vehicle passing event from which a vehicle signature can be extracted. This signature is unique for each vehicle, allowing us to implement a machine-learning algorithm for vehicle classification purposes. The system has been experimentally tested by performing measurements using dark fiber in a telecommunication fiber cable running in a buried conduit along 40 km of a road open to traffic. Excellent results were obtained, with a general classification rate of 97.7% for detecting vehicle passing events and 99.6% and 85.7% for specific car and truck passing events, respectively.

Scanning-EMG is an electrophysiological technique in which the electrical activity of the motor unit is recorded at multiple points along a corridor crossing the motor unit territory. Correct analysis of the scanning-EMG signal requires prior elimination of interference from nearby motor units. Although the traditional processing based on the median filtering is effective in removing such interference, it distorts the physiological waveform of the scanning-EMG signal. In this study, we describe a new scanning-EMG signal processing algorithm that preserves the physiological signal waveform while effectively removing interference from other motor units. To obtain a cleaned-up version of the scanning signal, the masked least-squares smoothing (MLSS) algorithm recalculates and replaces each sample value of the signal using a least-squares smoothing in the spatial dimension, taking into account the information of only those samples that are not contaminated with activity of other motor units. The performance of the new algorithm with simulated scanning-EMG signals is studied and compared with the performance of the median algorithm and tested with real scanning signals. Results show that the MLSS algorithm distorts the waveform of the scanning-EMG signal much less than the median algorithm (approximately 3.5 dB gain), being at the same time very effective at removing interference components.

In this thesis, several algorithmic procedures to estimate the MU fiber parameters from the waveform of a scanning-EMG signal (i.e., from the recording of a MUP in multiple positions along a linear corridor) have been developed. The different procedures correspond to different modifications of the scanning-EMG technique, which differ in the number of scanning needles and recording ports used to record the signal. The three proposed variants are: the 1-port recording, based on a single scanning needle with a single recording port placed at one side of the needle (i.e., a single fiber EMG needle); the 2-port recording, based on a single scanning needle, with two recording ports, placed at opposite sides of the needle; and the 4-port recording, based on two scanning needles, each one with two recording ports. The estimation system also uses a linear array of surface-EMG recordings to obtain complementary information about the MU. In this way, the recording setup to achieve the MU parameter estimation consists on a simultaneous surface- and scanning-EMG signal recording of the MUP. The estimation system has been evaluated for the three scanning-EMG recording configurations (1-port, 2-port, and 4-port), and compared to the case in which the estimation is performed from a MUP recorded at a single position. The evaluation has been done in a simulation framework, using state of the art models of the muscle, MUs, recruitment, and needles, and developing specific models for the simultaneous surface- and scanning-EMG recording process. This provides a controlled environment in which the performance of the system can be objectively quantified and evaluated.The results evidence that MU parameters are estimated much more accurately when using the scanning-EMG technique than when using a MUP recorded at a single position, corroborating the hypothesis that the use of signals recorded at multiple positions enhances the parameter estimation. Among the three proposed recording configurations, the poorest estimation results have been obtained for the 1-port configuration which, moreover, is only capable of estimating the MU fibers at one side of the needle. The 2-port configuration gives better results, and allows to estimate the MU fibers at both sides of the needle. The 4-port configuration is the one that provides the best performance, but it has the disadvantage of being the most difficult configuration to be physically implemented. In the view of these results, a deeper evaluation of the 2-port recording configuration is done. This is because it combines a good estimation performance with a relative ease to be physically implemented. An additional effort is done to calculate several global MU parameters, such as the MU fiber density, the average potential propagation velocity of the fibers, and the width of the innervation zone, from the resulting set of estimated fibers. These global MU parameters provide relevant physiological information from a clinical point of view. Hence global parameters connect the estimation system developed in this thesis with a future application in the diagnosis and follow-up of neuromuscular pathologies.

A novel processing and classification framework for the automatic detection and classification of pedestrians, bicycles, and cars in urban environments is proposed and experimentally demonstrated. The proposed processing pipeline combines high-sensitivity distributed acoustic sensing, automatic trajectory extraction, physically interpretable spectral feature extraction, and supervised machine-learning-based pattern recognition. The technique leverages pre-existing telecommunication fibers installed beneath or close to sidewalks and roadways. Coherent optical time-domain reflectometry enhanced with optical pulse compression provides the sensitivity and spatial resolution required to detect mobility-induced vibrations using these non-dedicated fibers, thereby avoiding sensing-cable deployment or pavement retrofitting. A detailed analysis of the signals extracted along the detected trajectories is performed in the spatial-frequency and equivalent-frequency domains. Controlled walking, running, and cycling measurements, together with passenger-vehicle observations under normal traffic conditions, reveal reproducible spectral characteristics associated with the underlying pavement-loading mechanisms. In particular, the analysis identifies four complementary properties: spatial loading periodicity, periodic loading peak prominence, initial spectral decay, and broadband spectral-energy distribution. Seven physically interpretable features derived from these properties are subsequently supplied to a support vector machine classifier. The processing and classification framework is evaluated using measurements acquired from an existing optical fiber in a real urban environment containing simultaneous sidewalk and road activity. When classification is restricted to manually verified bicycle, car, and pedestrian trajectories, a balanced accuracy of 98.3% is achieved. These results demonstrate the potential of physically interpretable spectral analysis of DAS measurements for continuous, spatially distributed, and privacy-preserving urban mobility monitoring.