Picture recognition and analysis

Picture recognition and analysis has numerous applications in all fields of technology, from medical diagnosis to document processing. This paper introduces some of the basic techniques of picture analysis: correlation, or matching of pictures; transformations and their uses, including their application to normalization for recognition purposes; parametrization, or property measurement, for picture classification; and segmentation of a picture into parts for purposes of description, with emphasis on segmentation based on region connectivity. Examples are given showing how each of these techniques is important in practical applications.