A neural network architecture is presented for monitoring events coded in image time sequences. The image sequence defines the sampling of high frequency phenomena where only morphological aspects of the scene are taken into account. In particular a model for implementing a hierarchical neural network architecture is proposed. Preliminary results are shown in the study of the oscillation states in the flame front of the power plant gas combustors.
A multilevel neural approach to dynamic scene analysis
Salvetti O
2003
Abstract
A neural network architecture is presented for monitoring events coded in image time sequences. The image sequence defines the sampling of high frequency phenomena where only morphological aspects of the scene are taken into account. In particular a model for implementing a hierarchical neural network architecture is proposed. Preliminary results are shown in the study of the oscillation states in the flame front of the power plant gas combustors.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
prod_68276-doc_123098.pdf
solo utenti autorizzati
Descrizione: A multilevel neural approach to dynamic scene analysis
Tipologia:
Versione Editoriale (PDF)
Dimensione
94.69 kB
Formato
Adobe PDF
|
94.69 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
