Region segmentation: Signal vs semantics

The problem of region segmentation includes issues varying from the preprocessing of the image to its semantic interpretation. Rather than surveying various research activities technique by technique (such as merge/split or semantic/nonsemantic), this paper tries to give a unified view of this problem. We first briefly present a model of image understanding to provide a paradigm in which the problem of region segmentation can be discussed. The model enables us to identify three levels of knowledge—signal, physical, and semantic—which play different roles in achieving the goals of region segmentation. For each level of knowledge, the development of research in that level is first reviewed, then its important issues are discussed. This paper emphasizes the importance of exploiting the physical level of knowledge, the bridge between a picture and a scene.