Mobile Intelligence

21.2 RELATED WORK

Video browsing and retrieval in mobile devices is an emerging research area. Due to the constraints of mobile devices in terms of their power consumption, processing speed, and display capability, more challenges have been encountered than in traditional multimedia applications and many research studies have been conducted to address various issues.

To reduce the viewing time and to minimize the amount of interaction and navigation processes, a variety of studies in academia and industry have been carried out on the summarization of video contents. For instance, in Ref. [11], singular value decomposition (SVD) of attribute matrix was proposed to reduce the redundancy of video segments and thus generate video summaries. Clustering techniques were also used to optimize key frame selection based on visual or motion features to enhance video summarization [2]. In industry, Virage has implemented preliminary video summarization systems for NHL hockey videos using multimodal features [29]. However, it remains a big issue in terms of the semantic gap between computable video features and the meaning of the content as perceived by the users. For this purpose, metadata about the content was utilized and played an active role in video retrieval [25]. For instance, ontologies have been proposed in Ref. [17] to perform intelligent queries and video summarization from metadata. In Ref. [28], a video semantic summarization system in the wireless/mobile environments was presented, ...