Advances in Road Feature Detection and Vehicle Control Schemes: A Review

Abstract

Road accidents are a major cause of fatalities globally. These accidents are caused by human errors, which include over speeding, drowsiness, intoxication, and loss of concentration. In an attempt to overcome these challenges, Autonomous Vehicles (AVs) have been developed and the prominence of these vehicles is rapidly growing. AVs are classified into five (5) steps ranging from no automation to full automation. For a successful operation of an AV, numerous features are put in place, one of which is Road Feature Detection and Vehicle Control. Several techniques have been adopted to ensure an effective feature detection in AVs. This paper presents a review and survey of existing techniques in lane detection and identification of road anomalies with respect to AVs. An overview of AVs and computer vision are presented, as well as the features, strengths, weaknesses of existing literature. Existing schemes were discovered to be unfit for unstructured and complex environments due to a lack of consideration for nonlinearities and an inability to perform in real-time scenarios.