Visual video retrieval system using MPEG-7 descriptors
Vojtěch Zavřel, Michal Batko, Pavel Zezula · 2010
Video is one of the most popular media these days. However, the volume of the video content grows very fast, e.g. about 24 hours of video content is uploaded every minute to the most famous web site YouTube, and the necessity to search in this content is apparent. Generally, most of the video search systems are based on annotations or use additional text information, which is not always of high quality or lack precision. We present a system that allows user to search videos according to their visual content.