It would be very beneficial to have a computer vision system that computes the 3d bounding box of an object. This is very different to make a 2d bounding box. Tecnhnically what we do right now is calculate the centroid of an object using the depth information of the camera, and bounding box creation using YOLO. We have to research for methods that do this bt making a bounding entirely 3D.
It would be very beneficial to have a computer vision system that computes the 3d bounding box of an object. This is very different to make a 2d bounding box. Tecnhnically what we do right now is calculate the centroid of an object using the depth information of the camera, and bounding box creation using YOLO. We have to research for methods that do this bt making a bounding entirely 3D.