Pavement detection in YCbCr color space and its application

Abstract

Recognition of driving-view image is the most important future issue in a smart tachograph. The intersection imaging recognition is one of the important issues in driving-view image recognition. In addition to using in the GPS satellite positioning correction in the metropolitan area, it can also be applied to the traffic lights imaging recognition in an intersection. With some samples analyzing, the color ranges of pavement in the YCbCr color space is developed in this paper. Therefore, pavement and markings on the road can be separated. By using this color separation and calculating the ratio of markings and pavement, it can be applied to intersection imaging recognition of driving-view image. Experiments are carried out for some VEDR's videos to demonstration the computational advantage of the proposed method.