A vector-correlative method for dynamic object state recognition

Author

Boriskevich, A.A. ; Khrabrov, V.V.

Author_Institution

Inst. of Eng. Cybern., Acad. of Sci., Minsk, Byelorussia

Volume

2

fYear

1997

fDate

2-4 Jul 1997

Firstpage

949

Abstract

A vector-correlative method to recognize the dynamic object state in multidimensional space based on the similarity measure in the form of a modified scalar product, the iterative training procedure and iterative adaptation to the set of recognizing samples defined by the expert are proposed. The method provides high sensitivity to small object state changes and stability of recognition results to the noise level change in the signal under study. The method allows to recognize the current object state for a minimal number (two or three) of iterations

Keywords

correlation methods; iterative methods; noise; pattern recognition; signal sampling; state estimation; dynamic object state recognition; expert; iterative adaptation; iterative training procedure; modified scalar product; multidimensional space; noise level change; similarity measure; small object state changes; stability; vector-correlative method; Character recognition; Cybernetics; Iterative methods; Mechanical engineering; Medical diagnostic imaging; Multidimensional systems; Noise level; Sampling methods; Signal processing; Stability;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Signal Processing Proceedings, 1997. DSP 97., 1997 13th International Conference on

Conference_Location

Santorini

Print_ISBN

0-7803-4137-6

Type

conf

DOI

10.1109/ICDSP.1997.628520

Filename

628520