1 September 2002 Bad pixel identification by means of principal components analysis
Author Affiliations +
Optical Engineering, 41(9), (2002). doi:10.1117/1.1497397
Bad pixels are defined as those pixels showing a temporal evolution of the signal different from the rest of the pixels of a given array. Principal component analysis helps us to understand the definition of a statistical distance associated with each pixels, and using this distance it is possible to identify those pixels labeled as bad pixels. The spatiality of a pixel is also calculated. An assumption about the normality of the distribution of the distances of the pixels is revised. Although the influence on the robustness of the identification algorithm is negligible, the definition of a parameter related with this nonnormality helps to identify those principal components and eigenimages responsible for the departure from a multinormal distribution. The method for identifying the bad pixels is successfully applied to a set of frames obtained from a CCD visible and a focal plane array (FPA) IR camera.
Jose Manuel Lopez-Alonso, Javier Alda, "Bad pixel identification by means of principal components analysis," Optical Engineering 41(9), (1 September 2002). https://doi.org/10.1117/1.1497397


Image quality testing of assembled IR camera modules
Proceedings of SPIE (October 25 2013)
Design And Performance Of 64 X 128 Element PtSi Schottky...
Proceedings of SPIE (December 28 1982)
PCA-based image fusion
Proceedings of SPIE (May 08 2006)
Infrared on-orbit RCC inspection system (IORIS)
Proceedings of SPIE (March 28 2005)

Back to Top