2D qualitative shape matching applied to ceramic mosaic assembly

Abstract

A theory of shape recognition of 2D objects and its application in the ceramic industry for intelligent automation of the mosaic mural assembly process are presented in this paper. This theory qualitatively describes the shapes of the objects, considering: (i) shape boundary characteristics, such as angles, relative length, concavities, and curvature; and (ii) their color and size. The shapes to be recognized may be regular or irregular closed polygons, or closed curvilinear figures. Each figure is described as a symbolic character string that contains all its distinctive characteristics. This description is used to determine whether the shape of two figures matches. Then, given a design of a mosaic and given a set of physical ceramic tesserae, an application is developed in order to recognize the tesserae that form the mosaic, thus enabling the intelligent and automated assembly of ceramic mosaics.