Invariant shape representation by Radon and wavelet transforms for complex inner shapes

This paper proposes a novel and effective invariant shape representation by Radon and stationary wavelet transforms for images with complex inner shapes. The proposed method is invariant to general geometrical transformations. Instead of analyzing shapes directly in the spatial domain, the proposed method retrieves features in Radon transform domain by statistical and spectral analysis to make shapes translation and size invariant. The stationary wavelet transform is used to make shapes rotation invariant. Experiment results show that the proposed shape representation is invariant to rotation, translation, and/or scaling changes for images with complex inner shapes, and outperforms the other existing methods with higher recognition rates.