FIELD: video encoding.
SUBSTANCE: method for dividing a group of pictures (GOP) is proposed, comprising the following steps: determining whether a scene switching frame is present in the current GOP of the received video data; determining whether the last video frame of the current GOP group is a scene switching frame; determining the first scene switching frame as the target scene switching frame of the current GOP; obtaining a target scene switching frame by viewing at least one scene switching frame in the current GOP from end to start; determining the previous video frame of the scene switching target frame as the last video frame of the current GOP; and determining a predetermined number of video frames that follow the scene switching target frame and include the scene switching target frame as the next GOP group and determining the next GOP group as the current GOP group.
EFFECT: prevention of error propagation caused by inter-frame referencing during encoding, which ensures high-quality referencing of subsequent video frames.
18 cl, 7 dwg
