Research On Static Multi-purpose Video Stitching

With the improvement of computer and camera performance, video acquisition and processing cost are reduced, video stitching gradually developed on the basis of image stitching. Video stitching technology has been used in monitoring system to obtain a big scene and high-resolution video. And video stitching have important and widespread application in our daily life, such as video conference, virtual reality, etc.In this paper the research based on multiple stationary camera video joining together of the same scene. First introduced the different images and video, and the process of image stitching is introduced in detail, also the process of SURF extraction is discussed, the method of coarse matching SURF and using brightness contrast operator accelerated coarse matching method, the improved RANSAC based similarity to calculate transformation matrix, image fusion, computing area is applied to the fusion of the next medium. Innovation point of this paper is to put forward an improved RANSAC algorithm based on similarity, this paper proposes a new method of computing image fusion area, and proposes a set of complete video fusion scheme, through the scheme can finish multi-purpose stationary video splicing in real-time, and ensure the accuracy of splice results.