| CPC H04N 19/132 (2014.11) [H04N 19/105 (2014.11); H04N 19/172 (2014.11); H04N 19/177 (2014.11)] | 15 Claims |

|
1. A method for processing a video frame, comprising:
acquiring a target characteristic of the video frame in a video frame sequence to be encoded, in a case of a quantity of video frames included in the video frame sequence to be encoded is greater than a preset threshold; wherein a type set of the target characteristic is a sub-set of a type set of a preset characteristic of a video frame sample used for training a first preset model to obtain a first target model, and the type set of the preset characteristic comprises at least one of:
an intra cost of a given frame,
a quantity of intra blocks of the given frame,
an intra cost of another frame associated with the given frame, or
a quantity of intra blocks of the other frame associated with the given frame;
inputting the target characteristic to the first target model, to obtain a first output result corresponding to the video frame; wherein the first target model is configured to predict the first output result based on the target characteristic, and the first output result indicates a first target group of pictures (GOP) length corresponding to the video frame; and
determining the first target GOP length, based on the first output result, comprising:
determining the first target GOP length based on the first output result and a mapping relationship between an output result and a GOP length established in advance; or
obtaining the first target GOP length based on the first output result directly indicating the first target GOP.
|