Abstract
This work presents a novel feature fusion method for texture retrieval. Considering the advantages of both the spatial and frequency domain, we first carry on the experiments in spatial domain and frequency domain respectively. On one hand, sober and histogram feature are used to calculate the similarity. On the other hand, Fourier is applied to obtain the frequency feature. Then a feature fusion scheme is used to join the two features came from spatial and frequency domain. Experimental results on MIT texture database show that the proposed method is effective.
Full Text
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call