A High-Level Survey of Video Annotation and Retrieval Systems

Robert Sorschag · 2012

Videos are an integral part of current information technologies and the web. The demand for efficient retrieval rises with the increasing number of videos, which is equally true for video annotation techniques as matadata is the primary source of most retrieval systems. Existing surveys about video annotation and retrieval cover specific approaches and the underlying analysis technologies. In contrast to these surveys, we focus on the big picture that is common for all of today’s systems, their distinguishing features, and we try to point out research trends and access points for further reading. This high-level survey addresses researchers that are new to the field and require a proper overview as well as individuals and enterprises that want to make up their minds about the annotation and retrieval possibilities they have. In this work, we start with a general definition of video documents before different aspects of annotation and retrieval systems are discussed in combination with user intentions and interaction processes. KeywordsContent-Based Video Analysis; Video Search and Retrieval; Metadata; Automatic and Interactive Annotation

Read the paper · More papers on PaperTik