Descriptor based on spectraln peaks correlograms
We have presented in this report a new peak descriptor SD based on the lag plot and discussed its relationship to the NBD. The SD, defined as the slope of a linear regression model constructed through the data samples belonging to a time sequence, shows to be a good candidate for describing sinusoi...
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Objeto de conferencia |
| Lenguaje: | Inglés |
| Publicado: |
2015
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/50546 |
| Aporte de: |
| Sumario: | We have presented in this report a new peak descriptor SD based on the lag plot and discussed its relationship to the NBD.
The SD, defined as the slope of a linear regression model constructed through the data samples belonging to a time sequence, shows to be a good candidate for describing sinusoidal and noise peak classes.
A proper choice of data set is crucial for discerning correctly between the peak classes. The SD makes use only of the spectral peak shape and consequently this information can be explained by just two parameters, namely the root mean square BWrms and absolute bandwidth L. This is similar to the way the shape factor explains the filter spectral shape in the circuit theory. We have shown that BWrms and L, initially defined in the spectrum domain, hold relationship to the time duration of the data set and its sampling rate respectively |
|---|