I have used the code in this repository linked below and implemented a lane segmentation system:
https://github.com/MaybeShewill-CV/lanenet-lane-detection
Then I found this demo which is Tesla Vision Path Prediction To See Around Blind Corners.
https://www.youtube.com/watch?v=rSAIWdcLPpo
i was kind of thinking that if I could be able to estimate slopes of the detected lanes and draw their tails in the image till they meet each other on intersection.
anyway, I wonder if anyone can help me to find any algorithm, paper or code that can use the detected lanes and output the predictions?
Thanks.