Video frame insertion processing method

Abstract

The invention relates to the technical field of image processing, in particular to a video frame insertion processing method, which comprises the following steps of: performing simple similarity comparison on two frames of images of an original video to judge whether a large-scale pixel change exists between the two frames of images so as to determine whether a background is changed, and if the large-scale pixel change does not exist in the images, supplementing frames by using the existing optical flow method; if the two frames of images exist, pushing the boundaries of the original images to the changed boundaries by extracting the boundaries before and after the change, and extracting a plurality of frame insertion images from the boundaries, so that frame supplementation is performed between the two frames of images. Optical flow information does not need to be collected, and meanwhile, a moving object in the images does not need to be recognized, so that the calculation amount is relatively small, and the application range is wide.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a video frame insertion processing method. Background technique [0002] In the case of unsatisfactory network conditions, users often need to actively lower the frame rate of the video in order to ensure the quality of the video image, and the video data is transmitted at a lower bit rate, so that the high resolution and high frame rate of the video cannot be satisfied at the same time, affecting the quality of the video Viewing effect, so it is necessary to interpolate the video frame to ensure clear and smooth video playback. The video frame interpolation technology in the prior art usually needs to perform motion estimation on the objects in the scene, and use the motion compensation algorithm to insert the objects into the correct position of the generated frame. Therefore, the effect of frame interpolation largely depends on the quality of motion estimation and com...

AI Technical Summary

Problems solved by technology

[0002] In the case of unsatisfactory network conditions, users often need to actively lower the frame rate of the video in order to ensure the quality of the video image, and the video data is transmitted at a lower bit rate, so that the high resolution and high frame rate of the video cannot be satisfied at the same time, affecting the quality of the video Viewing effect, so it is necessary to insert frames to the video to ensure clear and smooth playback of the video

The video frame interpolation technology in the prior art usually needs to perform motion estimation on the objects in the scene, and use the motion compensation algorithm to insert the objects into the correct position of the generated frame. Therefore, the effect of frame interpolation largely depends on the quality of motion estimation and compensation. The motion capture in the existing motion compensation algorithm is generally implemented by the optical flow algorithm, but the existing optical flow algorithm has a poor processing effect when the background changes and moving objects in the video exist at the same time

Images