IoT-Based Smart Alert System for Drowsy Driver Detection

Table 2

Performance analysis of the proposed system in view of 500-frame test in various conditions.

Face condition	TED	FED	TAD	FAD	TMD	FMD	Eye accuracy detection	Drowsy accuracy detection
Without glasses	498	2	497	3	495	5	99.6%	99.2%
With glasses	489	1	485	15	486	14	97.8%	97.1%
Alternative view	478	2	474	26	477	23	95.6%	95.1%
Finally	1465	5	1456	44	1458	42	97.7%	97.1%