A Novel Shape Descriptor Based on Interrelation Quadruplet

In this paper, we propose a new shape descriptor, which represents the 2-D shape information by using the concept of interrelation quadruplet. For this purpose, the polygonal approximation of 2-D shape is applied first. The line segments can be extracted from the polygonal shapes and the definition of interrelation quadruplet between ling segments is introduced. The properties of interrelation quadruplet that is invariant to translation, rotation and scaling of a pair of line segments is described. Several useful properties of the interrelation quadruplet are also derived in relation to efficient partial shape recognition. The shape recognition using the interrelation quadruplet requires only small space of storage and is shown to be computationally simple and efficient.