The research aims at finding out the following essential things.
1) Detecting multi-road lanes under complex weather conditions (fog, Snow, Dust, etc.)
2) Suppose road lane recognition can be enhanced with the road lane surface segmentation.
3) Increase lane detection reliability by using 3D detection methods/algorithms.
Prefer to :
1) Adding non-intervention, for instance (with no lane marking) as a comparison group.
2) Finding lane attributes such as lane types and other road elements such as crosswalks.
+ there is a custom dataset of 10 videos that needed to be labeled
+ writing 40-45 pages report about the work done and results.
Time : 25 days
Budget : 700-800$