Research On Video Key Frame Extraction Based On Motion Vector And Display In Virtual Reality Environment

In recent years,with the Internet video technology and video equipment developing fast,browsing video has become the main way to access information.In addition,due to the spread of smart phones,pads and a variety of video recording equipment,online video sites,live platform,video data sets and other video sources emerge rapidly and the amount of video data shows an explosive growth.The storage,sorting,screening and supervision of such a large amount of video data is an important problem that is urgently needed in Internet video technology.The amount of information carried by the video data is much larger than text and image data.Besides,video type tends to be diversified,and even the same type videos are different in content.The video viewer is difficult to obtain the information in a short time for such a large amount of video content.For users,there is a need for a way to quickly and accurately understand video content.However,the analysis and understanding of video content requires a lot of video information processing,which needs high cost and cannot guarantee real-time.Video key frame technology as a video summary technology,is very suitable for the above situation.In this paper,the algorithm of key frame extraction is introduced briefly,and a motion area detecting algorithm based on motion vector is proposed to improve the key frame extraction result.This method first calculates the color difference between adjacent two frames as the inter-frame difference by the relative entropy,and then uses the statistical tool to detect the outliers to obtain the shot boundary,thereby realizing the adaptive video shots detection.Based on shots detection result,motion vector is introduced.According to dividing the adjacent two frames into blocks and performing block matching,the sum of the block matching results is computed as motion vector.The magnitude of the motion vector reflects the degree of movement in the video,and therefore active and inactive area are obtained.Finally,key frames are extracted in these areas.In addition,this paper also studies the display of key frame extraction in virtual reality environment.Mainly through the change of information display mode and scene changes,experimental results are utilized to optimize the virtual reality environment for video key frame display mode.