Near-infrared video compositing

A video compositing system, comprising:

• a visible light source to emit visible light into a field of view in a first direction;
• an infrared light source to emit infrared light into the field of view in a second direction, the infrared light source included in a backdrop, wherein the first direction differs from the second direction;
• a dual camera system, including:
• a visible light camera, the visible light camera arranged to capture the visible light reflected from an object;
• an infrared camera, the infrared camera to capture the infrared light directed towards the dual camera system;
• a hot mirror disposed between the infrared camera and the visible light camera;
• wherein the hot mirror is adapted to filter infrared light from the field of view of the visible light camera, and reflect infrared light towards the field of view of the infrared camera;
• a video processing system adapted to receive visible video data from the visible light camera and infrared video data from the infrared camera, and to generate a video output;
• and
• a display screen arranged relative to the dual camera system, wherein the display screen reflects the video output provided from a display source while allowing the infrared light and the visible light to reach the dual camera system through the display screen.

A method for video compositing based on infrared video, comprising:

• capturing RGB video of visible light from a field of view using a visible light camera, wherein the visible light camera is arranged relative to a hot mirror, and wherein the hot mirror allows the visible light to travel through the hot mirror to the visible light camera;
• capturing infrared video of infrared light from the field of view using an infrared camera, wherein the visible light camera is arranged relative to the hot mirror, and wherein the hot mirror reflects the infrared light to the infrared camera, wherein the field of view is common to the visible light camera and the infrared camera;
• generating a matte of an object in the field of view from the RGB video captured by the visible light camera, based on removal of at least a portion of the RGB video using a silhouette of the object identified from the infrared video;
• and
• outputting a video including the matte to a display screen, wherein the display screen is arranged relative to the visible light camera and the infrared camera, to allow the infrared light and the visible light to reach the respective cameras through the display screen.