Video shot boundary detection using Petri-Net

Video shot boundary detection is an essential step for video browser and retrieval. Most shot detection methods detect cut translation without difficulty, but gradual translation is regarded as a hard problem. An approach for video shot boundary detection that can detect gradual transitions in addition to abrupt cuts is proposed. In our approach, the difference between two frames is measured by mutual information calculated separately for each of the HSV color components. Boundary frame of video shot is determined by the mutual information calculation. A Petri-net model is presented to describe boundary frames combination which indicates a shot boundary and used for video frames sequence to detect both cut transitions and gradual transitions. Experiments on test video data containing different types of shot transitions prove that our approach is efficient.