Abstract:
Autonomous driving perception should be environment-adaptive. Traffic line number and target system computing power should be considered in essential perception modules for traffic line detection. This paper proposes a traffic line detection method, Point Instance Network (PINet), based on key points estimation and instance segmentation. The PINet trains multiple hourglass models with the same loss function. The target environment’s computing power determines the trained model size. We use instance segmentation to train the PINet for clustering predicted key points regardless of traffic line number. On popular public lane detection datasets CULane and TuSimple, the PINet performs well.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
