Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Camera or Radar while turning on FSD

This site may earn commission on affiliate links.
While watching the FSD Beta I was wondering that, when the car makes a right turn, how would the car judge the speed of oncoming cars? It seems per the autopilot hardware graphics that the radar is only forward facing, and thus would not be able to know the speeds of oncoming cars during a right turn. So can the car judge the speed by camera alone? It seems that the left pillar camera would be best to see the oncoming cars.
 
While watching the FSD Beta I was wondering that, when the car makes a right turn, how would the car judge the speed of oncoming cars? It seems per the autopilot hardware graphics that the radar is only forward facing, and thus would not be able to know the speeds of oncoming cars during a right turn. So can the car judge the speed by camera alone? It seems that the left pillar camera would be best to see the oncoming cars.
The B-pillar cameras can see 80 meters ( 262 ft )
upload_2020-10-28_12-34-43.png
 
While watching the FSD Beta I was wondering that, when the car makes a right turn, how would the car judge the speed of oncoming cars? It seems per the autopilot hardware graphics that the radar is only forward facing, and thus would not be able to know the speeds of oncoming cars during a right turn. So can the car judge the speed by camera alone? It seems that the left pillar camera would be best to see the oncoming cars.
Presumably in a similar way to humans. No one really knows though since I'm sure it's part of the vision neural net.
Depth perception - Wikipedia
 
  • Like
Reactions: mikes_fsd
So can the car judge the speed by camera alone? It seems that the left pillar camera would be best to see the oncoming cars.
The entire purpose of the massive rewrite effort is to be able to take the camera feed and recreate the 3D + time birds-eye-view
The only way to recreate the 3D landscape is to be able to accurately determine the depth from the camera feeds - using cameras alone.