Flash is undergoing an explosive spread as a new prevailing media format on the Web. Unfortunately, few research efforts have been devoted to content-based Flash retrieval (CBFR) in IR community, which goes against the utilization of the enormous Flash resources. In this paper, Flash animation in 3-layer architecture is presented after it is segmented based on its visual features to a series of scenes on time-line of the production process. A promising approach of Flash animation retrieval is proposed based on the visual features of scenes and some meta-parameters. An experimental prototype system of Flash animation retrieval with roughly 100,000 Flash animations in total is built. The primary experiment demonstrates the flexibility and the effectiveness of our approach of CBFR. © 2008 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Meng, X., & Liu, L. (2008). On retrieval of flash animations based on visual features. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5093 LNCS, pp. 270–277). https://doi.org/10.1007/978-3-540-69736-7_29
Mendeley helps you to discover research relevant for your work.