ADAPTIVE APPROACH TO SIGNAL ANALYSIS (THROUGH THE EXAMPLE OF GEOACOUSTIC EMISSION)
Introduces the results obtained by the authors during the development of an integrated approach to the analysis of onedimensional geophysical signals. The approach is illustrated on geoacoustic signals. We illustrate the suggested approach using geoacoustic emission signals consisting of a sequence of relaxation impulses with a short front and long decay period. We discuss main approaches to constructing the feature space based on sparse representation of signal segments. Sparse representations are constructed using classical matching pursuit and its adaptive modification based on applying optimization techniques for iterative specification of the parameters of functions obtained as a result of signal decomposition. We provide the results of statistical analysis of decomposition results for real geoacoustic signals. We suggest an approach to segment classification based on extracted features.
Authors: A. B. Tristanov, O. O. Lukovenkova, D. M. Klionsky
Direction: Informatics and Computer Technologies
Keywords: Geoacoustic emission, signal data mining, sparse approximation
View full article