T. Tachikawa, K. Kameyama, T. Kawabe, K. Toraichi, N. Otsu, T. Kurita (Japan), and P.W.H. Kwan (Australia)
Content-based image retrieval, Smoothing, Image database,
Kamon, Function approximation.
A content-based retrieval method introducing smoothing is
proposed. The main aim of this work is to construct a
database of function approximated kamon images. By
introducing smoothing, kamon images can be characterized.
The proposed method can appropriately evaluate the
"similar" examples obtained by the questionnaire which is
difficult to match by the conventional method.