ABSTRACT
Analyzing a video shot requires identifying the meaningful components that influence the semantics of the scene through their behavioral and perceptual attributes. This chapter describes an unsupervised approach to identify meaningful components in the scene using perceptual grouping and perceptual prominence principles. The perceptual grouping approach is unique in the sense that it makes use of an organizational model that encapsulates
the criteria that govern the grouping process. The proposed grouping criteria rely on spatiotemporal consistency exhibited by emergent clusters of grouping primitives. The principle of perceptual prominence models the cognitive saliency of the subjects. Prominence is interpreted by formulating an appropriate model based on attributes which commonly influence human judgment about prominent subjects. The video shot is categorized based on the observations pertaining to the mise-en-sce`ne aspects of the depicted content.
