Format-Independent Motion Content Description based on Spatiotemporal Visual Sensitivity

With the extensive application of digital video technology, developing format-independent motion describing method is of great significance for retrieving and searching video content in different formats. In this paper, a format-independent motion describing method for video content is proposed. The features based on visual sensitivity are extracted from spatiotemporal slice. Since the same video content gives the same visual stimuli to visual perception, the method based on this kind of visual sensitivity related feature is format-independent. The experiments show the feature proposed is sensitive to the variation of video content and robust to the variation of video format. The motion describing method proposed is format-independent.