Share Email Print
cover

Proceedings Paper

Principal component analysis of noise in an image-acquisition system: bad pixel extraction
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The noise characterization of a set of frames can be treated by means of the principal component analysis. The main advantage of this method is that it provides a set of eigenimages that can be grouped into processes. These processes may be identified with actual sources of noise. In this scheme, bad pixels are extracted as those pixels showing an anomalous behaviour. The principal component analysis also allows to extract information about the character of the temporal evolution of the signal of the pixels. The bad pixels are identified by evaluating their place in the distribution of signal of the whole data set.

Paper Details

Date Published: 8 July 2003
PDF: 5 pages
Proc. SPIE 5036, Photonics, Devices, and Systems II, (8 July 2003); doi: 10.1117/12.498343
Show Author Affiliations
Jose Manuel Lopez-Alonso, Univ. Complutense de Madrid (Spain)
Javier Alda, Univ. Complutense de Madrid (Spain)


Published in SPIE Proceedings Vol. 5036:
Photonics, Devices, and Systems II
Miroslav Hrabovsky; Dagmar Senderakova; Pavel Tomanek, Editor(s)

© SPIE. Terms of Use
Back to Top