Video Stabilization Method Based on Image Stitching

Abstract

The invention provides a video image stabilization method based on image splicing. The method comprises: using an image stitching technology to splice videos captured by cameras into wide-angle view videos, performing center point detection on each frame of the spliced ​​video, and obtaining the original motion path of the camera through cumulative displacement vectors of the center points of adjacent frames; Fitting the original motion path with a fitting curve to obtain a smoothed motion path; intercepting an image from the video of the wide-angle view according to the smoothed motion path to regenerate a stable video. The method of the present invention aims at videos that can be spliced ​​into panoramas, uses the improved curve fitting method to obtain a smooth camera path, can effectively estimate the original motion trend of the camera, solves the disadvantage of large amount of calculation, and improves the algorithm efficiency; proposes Adjust the motion vector in the direction of the camera's own motion trend through iterative calculations to improve the smoothness of the track points and enhance the smooth visual experience.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a video image stabilization method based on image splicing. Background technique [0002] Vision is one of the important senses of human beings, and it is an important way for human beings to obtain various information. People perceive various information of the outside world mostly through vision. With the rapid development of modern information technology, video has become one of the important carriers of information, and has entered thousands of households, enriching people's daily life. People's requirements for video quality are getting higher and higher, and its uses are becoming more and more extensive, such as movies, TV, video chat, teleconferencing, video retrieval, etc., all need to process video signals. [0003] At the same time, electronic camera equipment has also become more and more common, ranging from large video cameras to small digital video cameras...

AI Technical Summary

Problems solved by technology

[0007] The disadvantages of the above-mentioned image stabilization method based on the two-dimensional motion model are: the feature trajectory in space is often sparse and unevenly distributed, and can start or end at any frame of the video; in addition, due to the presence of motion blur in the video taken by the user , texture area, camera translation and rotation, serious occlusion and other problems, long feature tracks are difficult to obtain

[0010] The disadvantage of the above-mentioned image stabilization method based on the 3D motion model is that 3D reconstruction is required. Using the SFM (Structure From Motion, structure from motion) algorithm to recover the structure from the motion requires a large amount of calculation, and in the absence of parallax, camera zoom, fast rotation, obvious Fails under conditions such as object movement and jelly effect

Images