Lane_F20

Lane detection has been gaining traction in terms of research focus for the past several years due to the rise in popularity of self-driving cars. Due to the autonomous nature of self-driving cars, their research and development naturally go hand in hand with lane detection research. Most recent works in lane detection research tackle the problem using machine learning algorithms to leverage their massively large data sets to train models that perform accurately under both favorable and unfavorable conditions. However, due to the long training time and the need for large data sets, our research aims to develop a method that can detect lanes quickly and relatively accurately under favorable conditions. Our implementation uses many of the key features of image processing and analysis. The key principle of this is that an image is a two-dimensional matrix of pixels with the value of each pixel representing the intensity of light there. Each pixel value ranges from 0 to 255 (inclusive). Our implementation uses a variety of images from the Jiqing Expressway Dataset to test the different features of our program. This project was written in Python. Please refer to the project paper for project details.