1 January 2004 Characterization of artifacts in fully digital image-acquisition systems: Application to web cameras
Author Affiliations +
Abstract
We apply principal component analysis (PCA) to the characterization of artifacts in a digital image-acquisition system containing image-compression algorithms. The method is successfully applied to web cameras. The classification done with the PCA method produces three processes. The pure spatial process retrieves the luminance distribution of a static object. The pure temporal process is directly related with the temporal noise of the system. An intermediate spatial-temporal process reveals the interaction between the compression algorithms and the spatial-frequency contents of the object. Without prior information, the PCA method is able to distinguish this interaction from the classical temporal noise. The analysis of the anomalous pixels also reveals the location in the scene where the compression algorithms work harder. An extension of this analysis identifies the origin of the anomalous behavior in terms of its spatial or temporal character.
© (2004) Society of Photo-Optical Instrumentation Engineers (SPIE)
José Manuel López-Alonso, José Manuel López-Alonso, Javier Alda, Javier Alda, "Characterization of artifacts in fully digital image-acquisition systems: Application to web cameras," Optical Engineering 43(1), (1 January 2004). https://doi.org/10.1117/1.1626667 . Submission:
JOURNAL ARTICLE
9 PAGES


SHARE
RELATED CONTENT

Quasi-Real Time Improvement In Small Format Images
Proceedings of SPIE (March 24 1976)
Spectral video intraframe compression and database
Proceedings of SPIE (January 16 2006)
High frame rate infrared and visible cameras for test range...
Proceedings of SPIE (September 07 1995)
Quasi-Real Time Improvement In Small Format Images
Proceedings of SPIE (August 31 1975)
Large fiber optic/CCD mosaic camera
Proceedings of SPIE (May 19 1993)

Back to Top