We propose an innovative, general purpose, approach to the selection and hierarchical representation of key frames of a video sequence for video summarization. In the first stage the shot detection module performs the video structural analysis; in the second stage the key frame extraction module creates the visual summary; and the last stage the summary post-processing module, after a pre-a classification aimed remove meaningless key frames, create a multilevel storyboard that the user may browse. Categories and Subject Descriptors H.3.1 [Information storage and retrieval]: Content Analysis xing – Abstracting methods, Indexing methods; I.4.10 [Image Processing and Computer Vision]: Image Representation – Multidimensional; I.5.3 [Pattern Recognition]: Applications – Computer Vision. General Terms Algorithms Keywords Dynamic video summaries, summary post-processing, supervised and unsupervised classification, multilevel visual summaries.