Computer analysis of long duration SEMG signals during simulated car driving: two different approaches

The aim of this paper is to detect the muscle fatigue using surface electromyography (SEMG) signal and then use the evolution of this fatigue to detect the best car seat. Two approaches are achieved and compared for two kinds of seats. The first approach is based on the calculus of the median frequency after detection of the SEMG segments. In this case we will compare the evolution of this median frequency for each seat. The second approach is based on the classification using wavelet transform and neural networks. The evolution of classes in each signal corresponding to each seat will gives information about the best car seat. These two approaches give the same result about the two seats.