Analysis of periodic signals via order statistic filters
Harold G. Longbotham, Thomas L. Arnow, Norman Barsalou, Jerome P. Keating, R.D. Glockman · 2002
The authors set forth an approach to the robust analysis of periodic digital data. This approach uses optimal order statistic (OS) filters for the filtering of non-Gaussian additive i.i.d. (independently identically distributed) noise and OS and generalized order statistic (GOS) filters to factor out impulsive and/or bursty noise. The noise is not necessarily stationary and may vary with phase (amplitude) and/or time. The methodology is applied to visual evoked response data.>