Signal detection, separation & classification under random noise background

We measured acoustic emission signals by means of Xedo-5 measuring device with the piezo-ceramic sensors. Subsequently we deal with the classification of acoustic emission random signals through model-based statistical clustering. The signals are compared by means of newly introduced attributes obtained directly from the signals and from normed frequency spectra such as φ-divergence distance measure originated from information theory and statistics (Hellinger divergence). We realize acoustic emission experiment to test the proposed classification attributes and further we perform a computer simulation study in order to find out the robustness and sensitivity of the corresponding signal classification techniques which can be applied to any other physical separation tasks such as particle collisions in high energy accelerator physics.