Automating assembly of ceramic mosaics using qualitative shape matching

This paper presents the intelligent and automatic application developed in order to assembly ceramic mosaics. The application consists on given an image with different objects (representing real tiles) and a vectorial image design of a ceramic tile mosaic, the application recognizes in a qualitative way which tile in the image belongs to the mosaic design and indicates its position and rotational angle in order that a robot arm places it in the correct position of the mosaic design. In order to implement this application two developments are needed and here presented: first a qualitative shape matching theory has been developed in order to be able to recognize shapes, and then a hybrid method (using quantitative and qualitative techniques) has been also developed in order to obtain the relevant features of an image. These relevant features are needed for their use in the qualitative shape matching theory. The qualitative shape matching theory considers qualitatively their angles, relative side length, concavities and convexities of the boundary and color. The shapes recognized are regular and non-regular closed polygons that can have curve segments and curvilinear shapes. Moreover the shapes can contain holes. Each object is described by a string containing its qualitative distinguish features, which is used to match the object against others.