Share Email Print

Proceedings Paper

Tensor invariant model for target discrimination
Author(s): Gianfranco Dacquino; Paolo Aschedamini; Rodolfo A. Fiorini; Antonio Meroni
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

A new physical, non-stochastic N-d model for target discrimination is presented. The model is based on Tensor Invariants and overrides usual stochastic procedure limitations problems characterized by FP and FN. The computational model is related directly to physical world, and it offers three major operational advantages over previous methods, at least. The first advantage is progressive automatic model generation of the Complete Minimum Set of Tensor Invariants. The second one is the reduced computational power requirements over traditional method. Finally, target precision drives automatic model generation trough subsequent steps. In fact, model precision is increased at each step. Robust discrimination or machine number representation saturation ends the computational process. Machine number representation saturation state suggests more power computational resource requirements for critical mission achievement. The general approach is tested on selected 2-D image database and preliminary results are presented.

Paper Details

Date Published: 6 August 2002
PDF: 11 pages
Proc. SPIE 4718, Targets and Backgrounds VIII: Characterization and Representation, (6 August 2002);
Show Author Affiliations
Gianfranco Dacquino, Politecnico di Milano (Italy)
Paolo Aschedamini, Politecnico di Milano (Italy)
Rodolfo A. Fiorini, Politecnico di Milano (Italy)
Antonio Meroni, Politecnico di Milano (Italy)

Published in SPIE Proceedings Vol. 4718:
Targets and Backgrounds VIII: Characterization and Representation
Wendell R. Watkins; Dieter Clement; William R. Reynolds, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?