Paper reading: Content Aware Video Compression:-An Approach To VOS Algorithm

Paper name	Content Aware Video Compression:-An Approach To VOS Algorithm
source	期刊 International Journal of Trend in Research and Development
years	2017
Author	Nisha Devi, Vinay Thakur, M.Tech
Core point	Use VOS algorithm and frame difference technology to realize video compression under H.264 coding standard.
Read date	2020.11.11
Impact factor
Number of pages	4
Number of citations
Content summary
The main problems and solutions the article solves: A content-aware-based video compression method is proposed to reduce the video size without reducing the video quality.   The main work of the article: The article proposes that the background in the video usually does not change position, while the foreground changes position. The displacement of the object in the frame can be determined by motion estimation, and only the displacement of one object relative to two objects is encoded. Motion estimation is a process of tracking small changes between consecutive frames to reduce redundancy between frames. After compressing the entire frame of a video, the new video is provided to H as input. 264 encoder. The H.264 encoder uses inter-frame encoding and compresses the entire video, so the synthesized video is stored in a limited storage space in a compressed form.   Article content: Algorithm implementation process: 1) Read the video frame and convert it into a separate frame; 2) Convert RGB image to grayscale image; 3) Set threshold to convert it to binary image; 4) Help in binary video image Next, extract the foreground object and subtract the background by using a filter to blur the background to obtain an adaptive background; 5) Pass the new segmented video to the H.264 encoder, and finally get the compressed video.   Experimental results: In the experiment, the author himself shot a 12-second video with a size of 11.7M. After compression, the size of the video is about 8.61M.