Step S1: performing feature extraction on a target frame in a network input to obtain a first feature, performing feature extraction on a first frame and an adjacent frame of the first frame and an optical flow between the first frame and the adjacent frame to obtain a second feature, and concatenating the first feature and the second feature to obtain a shallow layer feature, implementing initial feature fitting and adjusting a scale of a network;

Step S2: performing feature extraction and refinement on the shallow layer feature by using an iterative up and down sampling method to obtain a plurality of output first features and a plurality of output second features;

Step S3: performing feature decoding on the plurality of output second features, and concatenating decoded features along channel dimension to obtain features after a plurality of concatenation; and

Step S4: performing weight distribution on the features after the plurality of concatenation to obtain final features, and restoring an image by using the final features,

wherein the Step S1, Step S2, Step S3, and Step S4 effectively improve image quality.