The search functionality is under construction.
The search functionality is under construction.

Neuro-Fuzzy Recognition System for Detecting Wave Patterns Using Wavelet Coefficients

Sung Hoon JUNG, Doo Sung LEE

  • Full Text Views

    0

  • Cite this

Summary :

Recognition of specified wave patterns in one-dimensional signals is an important task in many application areas such as computer science, medical science, and geophysics. Many researchers have tried to automate this task with various techniques, recently the soft computing algorithms. This paper proposes a new neuro-fuzzy recognition system for detecting one-dimensional wave patterns using wavelet coefficients as features of the signals and evolution strategy as the training algorithm of the system. The neuro-fuzzy recognition system first trains the wavelet coefficients of the training wave patterns and then evaluates the degree of matching between test wave patterns and the training wave patterns. This system was applied to picking first arrival events in seismic data. Experimental results with three seismic data showed that the system was very successful in terms of learning speed and performances.

Publication
IEICE TRANSACTIONS on Information Vol.E84-D No.8 pp.1085-1093
Publication Date
2001/08/01
Publicized
Online ISSN
DOI
Type of Manuscript
PAPER
Category
Pattern Recognition

Authors

Keyword