Visual Animations Content Feature Analysis Based on Image Information Extraction

摘要

Visual animation is a popular media on the web. It consists of many media elements such as text, graphics, images and sound. Based on the structure of flash animation files, this paper analyzes the content features of visual animation and builds a platform for extracting content features. Creating an index library for image features of flash animation is part of the research. The model is the development of content-based animation retrieval system. The media objects are extract the dynamic features of the moving objects, such as the motion trajectories and movement patterns of the media objects in the animation. Based on the extraction of media objects, dynamic features and interaction feature, high-level semantic features such as interaction degree, animation tempo and animation category can also be automatically extracted in the future. As a result, the creation of animation can be deduced to some extent based on these high-level semantic features.