A signal-processing-based technique for P300 evoked potential detection with the applications into automated character recognition

EURASIP Journal on Advances in Signal Processing

Table 1 Listings of the P300 detection results on the 15-word testing dataset

Words for testing	M = 15	M = 14	M = 13	M = 12	M = 10	M = 8	M = 5	M = 3
CAT	CAT	CAT	CAT	CAT	CAT	CAN	CAN	ISN
FISH(1)	FISH	FISH	FISH	FISH	FISH	FISK	FISH	FIY H
FISH(2)	FISH	FC SH	FISH	FISH	FISV	FISH	FF SH	DAGJ
BOWL	BOWL	BOWL	BOWL	BOWL	BOWK	BOWK	BOXK	B1 WB
WATER	WATER	WATER	WATER	WATER	WATER	WATER	WATER	E5Z ER
HAT(1)	HAT	HAT	HY T	HAT	HAT	B AT	B AX	OYW
SHOES	SHOES	SHOES	SHOES	SHOES	SHOES	SHOES	Y HOES	YB1 ES
FOOD	FOOD	FOOD	FN OD	FN OD	FB OD	FN OD	FONB	FOH D
MOOT	MOOT	MOOT	MOOT	MOOT	MOOT	MOOT	MOOX	MOO4
HAM	HAM	HAM	HAM	HAA	HAM	HAM	B AA	HCQ
PIE	PIE	PIE	PIE	PIE	PIE	PIE	PIE	PIE
CAKE	CAKE	CAKE	CAKE	CAKE	CAKE	CAKE	CAKE	CAKC
TUNA	TUNA	TUNA	TUNA	TUNA	TO NA	TO NA	TO NA	TC NI
ZYGOT	ZYGOT	ZYA OT	ZYA OT	ZYA OT	ZYA OT	ZYA OT	ZYA OT	ZYA ON
4567	4567	4567	4567	4567	4567	458 7	X 58 7	R 58 7
Accuracy%	100.00	96.61	94.92	94.92	91.53	86.44	72.88	49.15

This dataset is for a variety of M values (results on experiment I). Note that each wrong detection character is set in italics.