Method and system for automatic television production
First Claim
1. A method comprising:
- capturing a scene comprising a sport event using a plurality of video cameras, to yield combined frame streams of video images covering together the sport event in its entirety;
analyzing the combined stream of video images to detect one or more objects associated with a playing object associated with the sport event;
deriving current and estimated position of the detected one or more objects and of the playing object;
generating a video production of the sport event by automatically selecting a sequence of portions of the combined stream of video images, wherein the selection is carried out based on predefined video production rules associated with a type of said sport event and further based on the current and estimated position of the detected one or more objects and the playing object;
estimating, upon losing said playing object by one of said objects, a region occupying said playing object in accordance with previous location thereof, andmodifying a frame stream of said one or more frame streams to include said region occupying said playing object.
1 Assignment
0 Petitions
Accused Products
Abstract
It is provided a method for a computerized, server autonomously producing a TV show of a sports game in a scene. The method includes receiving from several video cameras a stream of video images of the scene for capturing a panoramic view of the scene, analyzing the stream of video images for allowing definition of several frame streams, determining location data of the frame streams accordingly, and rendering an active frame stream with images imaging a respective portion of the panoramic view of the scene. The method includes also transmitting for broadcasting a stream of image frames imaging the respective portion of the panoramic view. The step of analyzing the stream of video images includes identifying a playing object, tracking the playing, object and identifying players. The method also includes calibrating the cameras using points in the playing field. The method may include analyzing the stream of video images for identifying an event in the scene for switching between the active frame stream and a different frame stream. Also, the method may include directing a directed sensor to a region of interest in accordance with location data of the active, frame stream.
-
Citations
9 Claims
-
1. A method comprising:
-
capturing a scene comprising a sport event using a plurality of video cameras, to yield combined frame streams of video images covering together the sport event in its entirety; analyzing the combined stream of video images to detect one or more objects associated with a playing object associated with the sport event; deriving current and estimated position of the detected one or more objects and of the playing object; generating a video production of the sport event by automatically selecting a sequence of portions of the combined stream of video images, wherein the selection is carried out based on predefined video production rules associated with a type of said sport event and further based on the current and estimated position of the detected one or more objects and the playing object; estimating, upon losing said playing object by one of said objects, a region occupying said playing object in accordance with previous location thereof, and modifying a frame stream of said one or more frame streams to include said region occupying said playing object. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
Specification