AUTHOR(S): 

Ayhan Küçükmanisa, Oguzhan Urhan

TITLE

Histogram Distance Based Road Surface Detection on a Single Image

KEYWORDS

Road surface detection, Lane detection, Quadratic-Chi histogram distance, ADAS

ABSTRACT

In this work, a novel histogram comparison based road surface detection approach which can be used in advanced driving assistance systems (ADAS) is proposed. The proposed method employs a lane detection approach to decide vanishing line. After the vanishing line is obtained, only the region below this point is processed. At the next stage, center of gravity (CoG) values of 1×k sized regions selected from the road surface is computed. This CoG computation is done using RGB histogram signature of the selected regions. The CoG slice is employed to check candidate road surface points by comparing histograms via Quadratic-Chi histogram distance. Experimental results show that the proposed approach is capable of detection road surface efficiently.

Cite this paper

Ayhan Küçükmanisa, Oguzhan Urhan. (2016) Histogram Distance Based Road Surface Detection on a Single Image . International Journal of Transportation Systems, 1, 84-91