Video Shot Boundary Detection Using Frame-Skipping Technique

Multimedia techniques have been used in many different appl ications over the past decade, sometime unmaturely. Demand for more efficient Efficient methods for video indexing and retrieval are required to confront this situation. Video temporal segmenta tion, aiming at finding the location of shot boundaries, including abrupt shot boundaries and gradual shot boundaries, is the first and essential step in visual data processing. In this paper we propose an algorithm for video shot boundary detection using frame-skippi ng technique. In our algorithm only some typical frames are selecte d for initial processing, which can be obtained by Frame-Skipped Mo ule (FSM). As a visual description, we may choose 1 or 2 frames from 10 consecutive ones. To diminish the error caused by Fra meSkipped technique, we use Abrupt-Boundary Module (ABM) and Gradual-Boundary Module (GBM) to detect video shots, where the ABM based on global histogram comparison is used to detect abrup t shot boundaries and the GBM based on local histogram comparison concentrates on detecting gradual transitions. This n ew algorithm can reduce about 40% 70% time-consumption compared with some previous algorithms on 4 typical video clips (movi e, cartoon, sports and news).