[Eeglablist] 1-Local binary pattern and number of neighbouring bits to be considered

[SAIRAMYA N J RR16EC001]

Sir,
     Currently, I am exploring the new techniques to classify epileptic and
non-epileptic signals. In which I came across a technique of using the
local binary pattern(Image processing) technique for 1-dimentional EEG
signal, in which they have incorporated the neighbouring bits to 8. As of
in image processing these techniques can be extended to 16 or 24
neighbouring pixels depending on the radius considered. For my EEG set I
obtain the best result when neighbouring bits is 12, is it possible to take
neighbouring bits as 12 in signal processing else the neighbouring bits
should be fixed to 8.




Thank you,
Ramya
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170720/7327d829/attachment.html>