Portfolio

Arena 3D

"Arena 3D" utilises four of our custom built AI cameras arranged by the user to mark out an arena (placing the 4x cameras in the 4 corners of the user's self defined arena). Activity within the defined arena is converted into a 3D scene. This is achieved by state of the art AI computer vision technology called '3D pose estimation' whereby the 4 cameras interpret 3D positions of key body points from 2D video. These marker points are then synchronised by the host device (a phone or tablet) which create a 3D scene from the interpreted footage.

The end result is similar to marker based MOCAP (motion capture) technology used in movies whereby actors wear special suits with reflective nodes that are interpreted into 3D positions. This existing MOCAP technology is hugely expensive, so our project here is exciting and upon completion, will bring this technology to new sectors who don't have the high budgets required for MOCAP technology.