Method and device for video frame synchronization

Abstract

The embodiment of the invention discloses a method and device for video frame synchronization. The method comprises the steps that two paths of obtained video are subjected to video framing processing, and an overlapped area in video frames in the two paths of video is solved; the two paths of video comprises first video and second video. One video frame which satisfies a first preset condition is chosen from the first video as a reference frame; according to a preset step length value, a video frame to be processed is chosen from the second video frame, and by calculating the similarity of the overlapped area of the reference frame and all video frames to be processed, a target video frame is chosen from the second video; according to the reference frame, the target video frame, a first frame rate of the first video and a second frame rate of the second video, the time difference value of the first video and the second video can be calculated; therefore, according to the time difference, correction for the video frame is completed. According to the method and device for the video frame synchronization, phenomena of blurring and 'ghosting' which occur during image splicing caused by time errors can be effectively avoided, and the accuracy of image splicing is improved.

Description

technical field [0001] The present invention relates to the technical field of video processing, in particular to a method and device for synchronizing video frames. Background technique [0002] Collaborative monitoring of large-scale video surveillance refers to overlapping the fields of view of two or more cameras to a certain extent; then registering the captured video image information; merging after obtaining the transformation relationship between the two images; thus The image information between different cameras is spliced ​​into large scene video image information. [0003] In the traditional way, researchers basically register and fuse the video images of different cameras by calibrating the cameras, so that the obtained large-scene video images can meet the requirements of low precision. However, because the traditional method ignores the different camera start-up time errors and the time error when the image sensor starts recording, the two frames of image inf...

AI Technical Summary

Problems solved by technology

However, because the traditional method ignores the different camera start-up time errors and the time error when the image sensor starts recording, the two frames of image information are directly stitched according to the static image stitching method, and the result obtained in this way is inaccurate.

Even when multiple cameras are started at the same time, there will be a certain time error between the first frames recorded by different cameras. If the time error is not considered, when there are moving objects in the overlapping area, blurring and "ghost" will appear during image stitching. shadow phenomenon

Images