Using simulated camera movements to create video from stationary images
The highly intuitive controller of the system is modeled on robotic cameras used by broadcasters, allowing the operator to simulate various camera movements on the stationary image, like panning sideways, tilting up/down, zooming in/out, and rolling. The shot memory function helps to create complex camerawork, and one can specify the duration of each shot down to a frame.
The system can handle photo data of up to 8,192 by 8,192 pixels and is capable of creating HD quality movies with zoomed-in images of up to 1,920 by 1,080 pixels.
The special effect feature includes adding shadows to or pixelating a stationary image. This will give producers room for more creativity than, for example, shooting a simple close-up of a paper document.
While continuing with our development to support 4K and 8K systems, we will also work on more sophisticated features such as recreating the subtle quiver of real-life cameras to offer a wider range of video creativity.
● This research is conducted in collaboration with Musashi Optical System Co., Ltd.
● This exhibit is presented by NHK’s Broadcast Engineering Department and NHK Niigata.