DocumentCode :
3045037
Title :
Pseudo-median filter with enhancing impulse suppression
Author :
Liu, Shaoqiang ; Chen, Lei ; Fan, Xiaoping ; Qu, Zhihua ; Yang, Xi
Author_Institution :
Sch. of Inf. Sci. & Eng., Central South Univ., Changsha, China
fYear :
2010
fDate :
20-23 June 2010
Firstpage :
1304
Lastpage :
1308
Abstract :
Pseudo-median filter is a computationally efficient alternative to median filter that applicable to data processing in low-level hardware platform of wireless sensor network nodes. In order to make the pseudo-median filter be able to remove impulses as does median filter and maintain the advantages of performance, a kind of improved pseudo-median filter algorithm with impulsive noise detecting and suppressing is proposed in this paper. By contrasting the increment of the maximum of the minima and the increment of the minimum of the maxima relative to the level of current filtering signal within the filtering window, the impulse can be detected and the output of signal polluted by the impulse is substituted by approximation. The simulation results and calculation complexity analysis show the improved algorithm obviously enhances the ability of impulse-attenuation and is a higher efficient alternative to median filter for the application in wireless sensor network.
Keywords :
filtering theory; impulse noise; median filters; wireless sensor networks; calculation complexity analysis; impulse suppression; impulsive noise detection; impulsive noise suppression; pseudo-median filter; signal filtering; wireless sensor network nodes; Adaptive filters; Band pass filters; Data processing; Finite impulse response filter; Hardware; Information filtering; Information filters; Nonlinear filters; Samarium; Wireless sensor networks; Impulse detecting and suppressing; Increment comparing; Median filter; Pseudo-median filter; Wireless sensor network;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information and Automation (ICIA), 2010 IEEE International Conference on
Conference_Location :
Harbin
Print_ISBN :
978-1-4244-5701-4
Type :
conf
DOI :
10.1109/ICINFA.2010.5512112
Filename :
5512112
Link To Document :
بازگشت