720-degree panorama video quick browsing method

Abstract

The invention discloses a 720-degree panorama video quick browsing method. A 720-degree panoramic video image is reconstructed by the back-projection method to obtain a corresponding view of the spherical viewpoint space in each direction of the line of sight. The lens length is determined by calculating the absolute brightness frame difference of adjacent image frames in a video sequence, and then a key frame is extracted. In this way, quick browsing of a panorama video is realized. By the invention, perspective views of a virtual scene in different view direction can be quickly generated and camera rotation and zoom movement are effectively simulated in each direction view, thereby improving the speed of browsing the virtual scene and well satisfying the specific application field of a virtual reality system.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to video panoramic image processing, in particular to a method for fast browsing of 720-degree panoramic video. Background technique [0002] With the development of information technology, people have higher and higher requirements for obtaining scene information with a wide range of viewing angles. Traditional photography methods can only obtain image frames with a limited range of viewing angles. Image stitching technology has emerged and developed rapidly to solve this problem. It combines two or more pictures with overlapping information into a complete image with an ultra-wide viewing angle to achieve the purpose of reducing image redundancy and obtaining wider viewing angle information. The generation of panoramic images is also an image stitching technology. A typical application of . [0003] 720-degree panoramic video is a video image sequence based on a spherical ...

AI Technical Summary

Problems solved by technology

[0006] One is a simple generation algorithm, which is a method of extracting key frames by uniformly sampling the video sequence at equal time, but this method is prone to the problem of excessive key frame extraction or insufficient representation due to the different changes in the amount of video information in a short period of time;

[0007] The second is the generation method based on visual information. According to the visual information such as color, shape, and texture in the video, various video processing technologies are applied to perform operations such as scene clustering, shot detection, and key frame extraction, and finally generate a thumbnail video. The method based on visual features has significantly improved the simple generation algorithm, but ignores the audio, subtitles and other information in the original video;

[0008] The third is to generate methods that integrate multiple features, such as using face recognition technology to detect the appearance of important people in news, using audio processing technology to detect highlights in sports videos, etc. feature fusion, the algorithm processing process is more complicated;

Images