| 摘要 |
There is described a method and system for recording, transmitting, and playing videos of an event-based format. Instead of providing a representation of each frame in a video as a complete image with brightness and color information for each pixel composing the image, each frame of a video is represented using a minimal data set that corresponds to an event that has taken place during the frame. Events are detected and recorded and the video may then be reconstructed by identifying actions that are triggered by the recorded events and executing the actions accordingly. Actions are identified by matching an event and an event coordinate, considering the objects and/or shapes present in a frame at the event coordinate and determining the consequences of the event on the objects/shapes.
|
