An Efficient Surveillance Video Coding Scheme for Static Camera Based Captured Video Data

摘要

Now-a-days video surveillance system is highly used in different organizations for security purposes. A video surveillance system based on static camera continuously captures video of a particular region. Due to the bulkiness of video data, main challenge is to efficiently store that huge data in a limited storage space. Conventional video compression algorithms are not that much efficient in terms of compression aspect. In this paper, we have proposed an efficient approach for surveillance video coding which is captured through static camera. Background subtraction technique is used to reduce temporal redundancy drastically between frames. An adaptive GoP length technique based on number of objects in a frame is introduced to get better compression of video data. Object based motion estimation technique will further reduce the video size in a conventional way. Our approach is tested on different captured video data and produces better results in terms of effective compression.