Autonomous cars in rural area roads

Abstract

Vehicles must be able to perceive their surroundings like human drivers in order to navigate roadways, wait at traffic lights and stop signals to avoid colliding with barriers like other vehicles and pedestrians. Focusing on the difficulties encountered by autonomous vehicles in recognising objects, a method has been developed to lane detection and tracking with the help of OpenCV library. The purposes and techniques towards using grayscale rather than colour, identifying edges in an image, choosing the zone of interest, implementing the Hough Transform, and using polar coordinates rather than Cartesian coordinates have all been addressed. The Canny Edge detection algorithm for lane detection is being implemented on a self-driving automobile prototype built on the Raspberry Pi. Our approach seeks to enable smooth lane detection for a quick and precise response to lane changes on the road. The implementation’s difficulties include assessing the road and lane borders, as well as predicting the proper degree of rotation of the wheel motor while keeping the vehicle's centre in line with the frame centre.