Abstract

One of the important roles of Driver Assistance System (DAS)is to assiststhe driver by alertingthe road anomalies. Detection of the speed bumps/breakers, potholes, and maintenance holes will fall under the road anomalies category.In India, generally there are two type of speed bump i) marking type speed bump ii) unmarking or non-marking speed bump. Among the two, using image processing technique identification of speed bump in earlier is comparatively easily than the later. Still, it is very challenging to detect the unmarked speed bump sincethere is a noyellow or white stripe to indicatetheir presence. During driving, the human visual system recognises the marking type speed bump in long-distance, but it is very tough in the case unmarked speed bump. So, a new method is proposed to identify the Real-time speed bump.The detection of such type of speed bump is essential for the driver to avoid accidents/inconvenience in driving. In the proposed method, after converting the RGB image into a grayscale image,a Gaussian filter is used to remove the noisy environment in the road image. Followed by canny edge detection to identify the edges of the image. This helps to locate the edges of speed bump though the colour of road and speed bump remains the same;there is a minute transition between them. Commonly the speed bump is constructed horizontally in the form of the linethat can be identified easy using Hough transform. On average, the accuracy ratio of corrected detected speed bump is 95.5%. In addition to driver assistance, the system can also be implemented in self-driving cars.