Eric Le Carpentier

Robustification of a map aided location process using road direction

Enhanced maps (Emaps) have proven their ability in aiding vehicle navigation, by applying constraints to lateral position of the vehicle. This is actually being demonstrated for instance in the frame of several lane level CVIS applications (Cooperative Vehicle Infrastructure Systems, a 6th Framework Programme Integrated Project). In some other research activities, the road direction between nodes or shape points of digital maps appeared to be of great interest as a possible way to improve vehicle heading estimation. Under this process, one assumes that both the vehicle and the road have most of time the same direction. In this article, we propose that the road geometry is described in Emaps by clothoids and contributes to vehicle location in a Bayesian framework by means of a particle filter. Actually, the road direction is a parameter very easily obtained when the road is precisely depicted by series of clothoids and acts as a strong constraint in the location process. Furthermore, the very challenging question of whether a GPS position should be accepted or rejected in such a filter has also been investigated. A strategy that aims at rewarding GPS instead of dead reckoning in case of lateral deviation is shown, with interesting results in particular for an on-board low-cost gyroscope.

Sequential decoding of intramuscular EMG signals via estimation of a Markov model.

This paper addresses the sequential decoding of intramuscular single-channel electromyographic (EMG) signals to extract the activity of individual motor neurons. A hidden Markov model is derived from the physiological generation of the EMG signal. The EMG signal is described as a sum of several action potentials (wavelet) trains, embedded in noise. For each train, the time interval between wavelets is modeled by a process that parameters are linked to the muscular activity. The parameters of this process are estimated sequentially by a Bayes filter, along with the firing instants. The method was tested on some simulated signals and an experimental one, from which the rates of detection and classification of action potentials were above 95% with respect to the reference decomposition. The method works sequentially in time, and is the first to address the problem of intramuscular EMG decomposition online. It has potential applications for man-machine interfacing based on motor neuron activities.

Online Intramuscular EMG Decomposition with Varying Number of Active Motor Units

This paper deals with the online decomposition of intramuscular electromyographic (iEMG) signals. A Markov model is proposed, which takes into account a varying number of firing motor neurons. A Bayes filter detects online the firing motor units by using a dictionary of approximated motor unit action potentials waveforms, and estimates precisely the action potential shapes and the respective firing rates. The method was tested on both simulated and experimental signals.