alimulfazalnabil/Autonomous-Lane-Detection ? reverse-engineered prompt

Build me a Python project for autonomous lane detection that works first as a simple OpenCV based road scene pipeline, and can also plug into CARLA for simulation later.

I want it to take an RGB road image, find the lane line candidates, estimate the left and right lane boundaries, smooth the result a bit over time, and save or show an annotated image so I can inspect what it found. Please keep the simulator part separate so the core detector still works without CARLA, but make it easy to run a CARLA client when the simulator is available.

Include a small test setup with synthetic lane images, some basic config defaults, and a simple example I can run right away with a sample road image. If you need to check current docs for OpenCV or CARLA integration details, go ahead and look them up online.