How it works

The figure below gives an overview of the iview system. The capture uses a number of time synchronised cameras. The minimal number of cameras is about four, but for good quality results a higher number is required. We are considering different configurations using broadcast coverage cameras and additional cameras.

The processing module computes a 3Dmodel of the scene. This is done using segmentation of objects from the background and 3D reconstruction. The replay module renders the captured scene in real-time using the computed 3D model and the original camera images deploying view-dependent texture mapping.

The entire system can potentially operate in real-time. At the current stage the processing is done offline. That means the images are stored and the processing is run at a later stage. The replay module is designed to work at interactive rates.

Here are some examples of rendered viewpoints:-

For more technical details please have a look into the published technical papers.