Key Frame Extraction from Video

Abstract

A complete overview of key frame extraction techniques has been provided. It has been found out that such techniques usually have three phases, namely shot boundary detection as a pre-processing phase, main phase of key frame detection, where visual, structural, audio and textual features are extracted from each frame, then processed and analyzed with artificial intelligence methods, and the last post-processing phase lies in removal of duplicates if they occur in the resulting sequence of key frames. Estimation techniques and available test video collections have been also observed. At the end, conclusions concerning drawbacks of the examined procedure and basic tendencies of its development have been marked.

Representative Based Clustering of Long Multivariate Sequences with Different Lengths

“…It should also be noted that advanced tools for film trailers making, generating of TV programs previews and other similar treatments to quick overview of the content can not be simple reductions of the original. Video processing requires to keep, at least, the restricted inquiries content aspects and to reduce the semantic gap between the features extracted from images (or/and their time sequences) [3,4].…”

“…Clustering of time series data induced by video streams, first and foremost, is aimed to an exploration of video tagging. Shots and key frames, as a rule, are adapted as units for excerpts labeling when segmenting video and searching for sets of similar temporal fragments, desirably, with nearly equivalent content [4,9]. If any known clustering algorithms can be applied when using key frames (at least in feature spaces) [10,11], then the clustering of time series is associated with a number of difficulties , the main one of which lies in different lengths of the sequences to be processed [12].…”