Algorithm for Autonomous Vehicle Control

Autonomous vehicle is one of the most researched topics in the current world. Accident prediction and avoidance is a major part of autonomous vehicle control logic. The control algorithm should have the ability to detect objects on the road, avoid accidents and travel in the required direction.The project provides autonomous vehicle control logic to video inputs captured through wide angle lens on a moving camera. The steps to be performed include identifying the road region from the whole frame by colour based segmentation and scanning, extracting lane markings and obstacles by scanning, calculating distance from the vehicle by mapping, determining available space on the sides of the vehicle.Distance and available space is used to determine the braking and steering angle change to avoid accidents. Here, outputs are shown for the whole video analysis through graphs indicating the distance to the obstacle, steering angle and breaking effect.