A Qualitative Theory for Shape Representation and Matching for Design

A new method for qualitative shape recognition and matching of objects in designs is presented in this paper. The approach consists of a reference-points information approach of the qualitative description of shapes considering qualitatively their angles, relative side length, concavities and convexities, and types of curvatures of the boundary. The shapes recognised are regular and non-regular closed polygons that can have curve segments and curvilinear shapes. Moreover the shapes can contain holes. To describe shapes with holes, topological and qualitative spatial orientation aspects are considered in order to relate the hole with its container. Each object is described by a string containing its qualitative distinguished features (symbolic representation), which is used to match the object against others. The paper also describes how this method can be used in industrial design by explaining an application. Given a file with different objects (representing tiles) and a vectorial image design of a ceramic tile mosaic border, the application recognises which tile in the file belongs to the border design and indicates its position and rotational angle to place the tile in the correct position of the border design. This qualitative method provides several advantages over traditional quantitative representations. The main advantages are the reduction of computational costs and the managing of uncertainty (two manufactured tiles are not geometrically identical, but they represent the same tile in the design).