The challenge of achieving vision in autonomous vehicles through traditional processing has been a topic of development for numerous years. To improve accuracy in depth estimation, expensive sensing devices have been introduced. This project aims to perform vision-related tasks, including segmentation and lane identification, using only a conventional monochrome stereoscopic camera and a Raspberry Pi 3b+ computing device. By executing all relevant processing operations directly on the Raspberry Pi, this approach demonstrates that accurate visual perception for autonomous vehicles can potentially be achieved using affordable and accessible computing hardware.