Shape Descriptor based on Invariant Generalised Ridgelet Transform for M x N Images

M.R. Mustaffa, F. Ahmad, R. Mahmod, and S. Doraisamy (Malaysia)

Keywords

Ridgelet transform, shape descriptor, RST invariant, arbitrary image size, precision and recall.

Abstract

Ridgelet transform has gained its popularity due to its capability in dealing with line singularities effectively at several levels of detail. This paper presents a novel approach for describing shapes based on Ridgelet transform. We propose a new implementation of the transformation called Invariant Generalised Ridgelet transform, which is invariant to rotation, scaling, and translation (RST) as well as able to handle images of arbitrary size. We introduce the implementation of Ridgelet transform on the ellipse containing the shape and the normalisation of the Radon transform. The 1D Wavelet transform is then applied to the Radon slices. In order to extract rotation invariant feature, Fourier transform is implemented in the Ridgelet domain. We test the new method on a standard MPEG-7 CE-1 B dataset in terms of precision and recall. From the experiments, it is shown that retrieval effectiveness has been significantly improved by 83.12%.

Important Links:

Go Back