REAL TIME DROWSINESS DETECTION USING DIFFERENT EDGE DETECTION TECHNIQUES

Gururaja A K, K.V.Mahendra Prashanth

Abstract: Irrespective of the advances in automobiles and roadway development a substantial extent of road accidents exist still because of driver drowsiness. Edge detection is a vital preprocessing image processing and computer vision technique. In the paper, a nonintrusive vision based frame work is employed for identification of car driver tiredness. Real time images have been considered for the experiments. Various edge detection techniques such as Sobel edge, Canny edge, Prewitt’s and Robert’s edge detection technique have been employed to detect the edges of the eye and determine the number of times the eye closes and opens. The results reveals that Sobel edge detection algorithm is easy to design as compared to other edge detection techniques. The output of Sobel algorithm demonstrates the proposed algorithm is more effective for edge recognition than other edge detection algorithms. The objective of the work carried out is design and development of a drowsiness detection system with more accuracy to reduce the incidents of road accidents.

Keywords: Drowsiness Detection, Edge Detection; Sobel Operator, Canny, Prewitt and Robert Operators.

DOI: https://doi.org/10.15623/ijret.2016.0506007