Self-Driving Car

Abstract

One of the major global challenges today is the huge number of traffic fatalities and the growing issue of traffic congestion. Self-driving cars are an advanced application of artificial intelligence, computer vision, image processing and embedded systems to enable autonomous navigation.

Self-driving cars are a smart combination of computer vision, programming, and robotics. In this project, we built a simple and low-cost self-driving car that can follow road lanes on its own. The system uses a Raspberry Pi and a USB webcam to capture live images of the road. The Raspberry Pi runs a Python program that uses the Open CV library to find lane lines and control the car’s movement.

The image from the camera is processed using basic image processing steps. These include converting the image to grayscale, removing noise, detecting edges, and focusing only on the region where lane lines are likely to appear. Once the lane lines are found, the Raspberry Pi sends signals to the motor using GPIO pins to keep the car inside the lane.

This project works only on lane tracking and does not include obstacle detection or traffic signal recognition. It is designed to be simple, affordable. The main goal is to show that even without expensive sensors or hardware a working self-driving system can be built using image processing and basic electronics.

This prototype is a good starting point for learning and future improvements. New features like obstacle detection or GPS navigation can be added later. The project proves that with basic tools and coding, it is possible to create a real-time, lane-following self-driving car.