I. Radev (USA) and G. Paschos (Greece)
Video Segmentation, Color Image Processing, Texture Anal ysis, Content-Based Retrieval, Film Modeling, Histograms.
This paper presents an integrated system for video annota tion according to a specific underlying video model. The sys tem is divided into three main components, namely a) video segmentation, where a video is separated into its constituent shots and representative frames (key frames) are collected; b) feature extraction, where features such as chromaticity mo ments and texture granularity are extracted from each key frame; and c) querying, where a user may select one of the collected key frames, and other key frames are retrieved based on similarity of the corresponding feature values. Experimen tal results are presented.
Important Links:
Go Back
