Video super-resolution method and apparatus

An electronic device, comprising:

• a memory and a processor,
• the memory being configured to store instructions;
• and
• the processor being configured to, when executing the instructions, cause the electronic device to implement a video super-resolution method, comprising:
• decomposing a t-th image frame of a video to be super-resolved and a neighborhood image frame of the t-th image frame, into N image blocks, respectively, both t and N being positive integers;
• generating N image block sequences according to the image blocks obtained by decomposing the t-th image frame and the neighborhood image frame, image blocks in the image block sequence being located at same positions of different image frames, respectively;
• calculating motion parameters of the image block sequences, motion parameters of any image block sequence being configured for representing optical flows between image blocks of adjacent image frames in the image block sequence;
• determining super-resolution network models corresponding to the image block sequences according to the motion parameters of the image block sequences;
• performing super-resolution on the image blocks of the t-th image frame in the image block sequences by using the super-resolution network models corresponding to the image block sequences, to obtain super-resolution image blocks of the t-th image frame;
• and
• generating a super-resolution image frame of the t-th image frame according to the super-resolution image blocks of the t-th image frame.

A non-transitory computer-readable storage medium having thereon stored instructions which, when executed by a processor, implement a video super-resolution method, comprising:

• decomposing a t-th image frame of a video to be super-resolved and a neighborhood image frame of the t-th image frame, into N image blocks, respectively, both t and N being positive integers;
• generating N image block sequences according to the image blocks obtained by decomposing the t-th image frame and the neighborhood image frame, image blocks in the image block sequence being located at same positions of different image frames, respectively;
• calculating motion parameters of the image block sequences, motion parameters of any image block sequence being configured for representing optical flows between image blocks of adjacent image frames in the image block sequence;
• determining super-resolution network models corresponding to the image block sequences according to the motion parameters of the image block sequences;
• performing super-resolution on the image blocks of the t-th image frame in the image block sequences by using the super-resolution network models corresponding to the image block sequences, to obtain super-resolution image blocks of the t-th image frame;
• and
• generating a super-resolution image frame of the t-th image frame according to the super-resolution image blocks of the t-th image frame.