An Optimised Deep Learning Approach of Lane Detection in Complex Indian Environment
Manas Kumar Rath, Prasanta Kumar Swain, Shobhan Banerjee
Abstract
Lane detection systems are useful to avoid road accidents that are due to wrong-lane driving. Such systems have the target to detect the lanes and warn the driver in case the vehicle enters the wrong lane. In this paper, we've proposed a lane detection technique that increases the accuracy and stability of the vehicle while tracing the position of the lane in a complex driving environment. Based on the existing methods like Canny edge detection and the Hough transform method, the proposed technique has an excellent detection ability and can detect key lanes in complex road environments such as darkness, traffic jams, and uneven lighting.
Topics & Concepts
Hough transformComputer scienceComputer visionTracingArtificial intelligenceCanny edge detectorLane departure warning systemEnhanced Data Rates for GSM EvolutionEdge detectionKey (lock)Stability (learning theory)Image processingImage (mathematics)Machine learningComputer securityOperating systemAutonomous Vehicle Technology and SafetyRobotic Path Planning AlgorithmsSmart Agriculture and AI