Share Email Print
cover

Proceedings Paper

The ghosting artifact reduction in high-order statistics based nonuniformity correction
Author(s): Wenfang Sun; Yigong Zhao; Hong Zhu; Beibei Song
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

This paper presents a nonuniformity correction (NUC) method based on high-order statistics to reduce "ghosting" artifact. Firstly, estimate the motion shift between the current frame and the previous frame of video sequences. When the motion shift exceeds the given threshold, we choose exponential windowing (EW) NUC method or else choose iterative high-order statistics NUC method. The strength of our method lies in its simplicity, low computational complexity, and its good trade-off between nonuniformity correction and "ghosting" artifact reduction. The ability of proposed method is demonstrated by using IR video sequences with both real and simulated nonuniformity.

Paper Details

Date Published: 4 March 2008
PDF: 10 pages
Proc. SPIE 6621, International Symposium on Photoelectronic Detection and Imaging 2007: Photoelectronic Imaging and Detection, 66211F (4 March 2008); doi: 10.1117/12.790836
Show Author Affiliations
Wenfang Sun, Xidian Univ. (China)
Yigong Zhao, Xidian Univ. (China)
Hong Zhu, Xidian Univ. (China)
Beibei Song, Xidian Univ. (China)


Published in SPIE Proceedings Vol. 6621:
International Symposium on Photoelectronic Detection and Imaging 2007: Photoelectronic Imaging and Detection

© SPIE. Terms of Use
Back to Top