Fourth International Conference on Advances In Computing, Control And Networking - ACCN 2016
Author(s) : CH.SATYANARAYANA , M. RADHIKA MANI, POTUKUCHI D.M.
Automated object recognition methods are essential for numerous applications of machine vision and pattern recognition. For an efficient object representation, a contour-based shape descriptor is designed, with a one dimensional shape signature. The present paper proposes a novel shape signature for recognizing the objects in complex plane. The proposed shape signature is applied on the contour shape representation, and then the description of representative shape features with the corner potential flow measure followed by the Fourier transformation. During recognition process, the Euclidean distance measure is evaluated to estimate the similarity score between the objects. The recital of the proposed shape descriptor has been checked with the Kimia 99 database. The experimental results are found to be robust and invariant to transformations.