A novel attention model and its application in video analysis

For the purpose of content based video process, this paper presents a novel video attention model. The impacts of the spatial position, the clearness, the appearance probability and motion vector on video attention regions are taken into account. Based on these extracted information, a novel model which can be employed to find video attention regions is finally constructed. Subsequently, after global motion estimation, an adaptive flag parameter of the camera motion is defined to apply the proposed model to videos with a still as well as a moving camera. Then an adaptive threshold choice is given for extracting video attention regions automatically, and attention curve of videos based on the definition of frame attention is given, which can describe the attention variation of viewers when video is showing. The promising results indicate the efficiency and accuracy of this model.