3 ms·
It isn't machine learning, it's aerial photography shot at an angle with depth sensing. For example, take an airplane and mount two cameras on it, one on each
by Stratoscope 8y ago
It isn't machine learning, it's aerial photography shot at an angle with depth sensing.
For example, take an airplane and mount two cameras on it, one on each side, angled down at perhaps 45 degrees to capture ground imagery in each direction.
Add a LIDAR device next to each camera to capture a "point cloud" of the same area the camera is imaging.
Now you have photographic images with distance data for each pixel. You can use that to construct a 3D image that can be viewed from various angles.
(Disclosure: I work at an unrelated Alphabet company but have no personal knowledge of any of this, it's just my semi-educated guess.)
- jakobegger 8y agoYou probably don't even need the LIDAR. Since the airplane is moving, you can use the parallax between successive images to determine depth information. The technique is called "structure from motion"
- jabberthemutt 8y agoOr include the position and orientation of the images' cameras and do a traditional photogrammetric bundle adjustment.
- Fiaxhs 8y agoI don't think they're using lidar, but I don't know much about it. However your comment made me think of this awesome article about Google maps and 3d buildings https://www.justinobeirne.com/google-maps-moat https://www.justinobeirne.com/google-maps-moat
- fleetingmoments 8y agoI thought it had to be accomplished by satellite somehow. I looked at a very remote island in the southern Indian Ocean where I was stationed and the amount of detail in Bing Bird's eye view of the geography was astonishing. This was circa 2013. I would be very surprised if they flew a plane out there just for those islands, there are no flights or shipping routes passing by there.
- thinkingemote 8y agoI think your reply had the clue. "Where I was stationed". Seems like it was valuable and valuable places get attention.