Project outline
A 3D visualization of people, directions and objects with the use of multiple surveillance cameras
- Detects people, objects and vehicles that show (divergent) directions and behaviors
- Is capable to calibrate and calculate camera images from different angles and transform it into a 3D visualization
- The visualization will give an accurate realtime location of people, vehicles and objects
Description
The 3D camera surveillance demo shows that we can combine multiple camera feeds and therefore calculate the 3D position of people and objects. With the use of deep learning algorithms we are capable of detecting people, objects and faces. We can even detect the direction someone is looking at. Because we calibrate different cameras at the same time we can calculate the detection and make it into a 3D coordinate system, which we then can visualize on to a 3D representation. With this technique it is possible to make an advanced spatial detection and follow objects whilst using normal cameras. The technique can also be used to define the distance between people or to detect divergent behavior of vehicles.
Description
The 3D camera surveillance demo shows that we can combine multiple camera feeds and therefore calculate the 3D position of people and objects. With the use of deep learning algorithms we are capable of detecting people, objects and faces. We can even detect the direction someone is looking at. Because we calibrate different cameras at the same time we can calculate the detection and make it into a 3D coordinate system, which we then can visualize on to a 3D representation. With this technique it is possible to make an advanced spatial detection and follow objects whilst using normal cameras. The technique can also be used to define the distance between people or to detect divergent behavior of vehicles.
