The Community for Technology Leaders
Multimedia Computing and Systems, International Conference on (1996)
Hiroshima, Japan
June 17, 1996 to June 23, 1996
ISBN: 0-8186-7819-4
pp: 0201
Atushi Ono , Information Systems Product Development Lab., SHARP Co. Institute of Industrial Science, Yarnatokooriyama, Nara
Masashi Amano , Information Systems Product Development Lab., SHARP Co. Institute of Industrial Science, Yarnatokooriyama, Nara
Mitsuhiro Hakaridani , Information Systems Product Development Lab., SHARP Co. Institute of Industrial Science, Yarnatokooriyama, Nara
Takashi Satoh , Univ. of Tokyo, MinaLto-ku, Tokyo
Masao Sakauchi , Univ. of Tokyo, MinaLto-ku, Tokyo
ABSTRACT
In this paper, we propose the image database with fully automated keyword extraction function. Our approach can extract two different conceptual level keywords from images. One is "Conceptual Keyword'', as we call it. Another is ``Scene Description Keyword''. "Conceptual Keyword'' is extracted by image recognition technique using "State Transition Model'', which is a hierarchical model. "Scene Description Keyword'' is extracted by primitive parameters of color segments. And we propose introduction of "Transition Probability'' to raise the retrieval accuracy (precision). Moreover, we evaluate the retrieval accuracy of this image database through a retrieval experiment using about 170 scenery images.
INDEX TERMS
State Transition Model, Conceptual Keyword, Scene Description Keyword, Transition Probability
CITATION
Atushi Ono, Masashi Amano, Mitsuhiro Hakaridani, Takashi Satoh, Masao Sakauchi, "A Flexible Content-based Image Retrieval System with Combined Scene Description Keyword", Multimedia Computing and Systems, International Conference on, vol. 00, no. , pp. 0201, 1996, doi:10.1109/MMCS.1996.534975
93 ms
(Ver 3.3 (11022016))