- Joined
- Jan 21, 2010
- Messages
- 25,510
one option is a pair of cameras and some image processing to locate the matching items in each. The displacement in the image can be used to calculate an angular separation and knowing the distance between the cameras will give you a distance. A GPS will give your motion vector. Comparing the relative position over time with your motion vector will allow you to determine if the other object is stationary. The next task is to predict the future positions of all the objects using their historical motion vectors and then (presumably) look for a conflict.
you may want to supplement this with proximity detection for very close objects that may look very dissimilar to the cameras, or as a method of detecting things in a "no go" zone.
So a basic system would need relatively simple sensors and some very clever (and fast) software.
you may want to supplement this with proximity detection for very close objects that may look very dissimilar to the cameras, or as a method of detecting things in a "no go" zone.
So a basic system would need relatively simple sensors and some very clever (and fast) software.