Detecting Obstacles Within the Driving Lane and Vehicle Speed Adjustment

Abstract

Driving lane detection and collision avoidance are common Advanced Driving Assistant Systems in modern vehicles. In this paper the algorithm for obstacle detection inside driving lane and vehicle speed adjustment is proposed. The proposed algorithm processes an image from a front view camera using traditional computer vision algorithms such as color filtering, Canny edge detection and Random sample consensus (RANSAC) to detect driving lane and to detect possible obstacle inside the driving lane. The abovementioned algorithms are implemented in the C++ programming language using a OpenCV library. The estimation of the obstacle distance from vehicle as well as automatic braking is based on data obtained using Carla simulator. The evaluation of the proposed algorithm is performed on still images taken from Carla simulator to show the effectiveness of obstacle detection and distance estimation while driving simulations are performed in order to evaluate vehicle speed adjustment in case of an obstacle within the driving lane.