Title :
A New Boundary Descriptor Based on the Directional Distance Histogram
Author :
Kong, Xiaodong ; Luo, Qingshan ; Guo, Ying ; Zeng, Guihua
Author_Institution :
Electron. Eng. Dept., Shanghai Jiao Tong Univ.
Abstract :
A new boundary descriptor for shape-based image retrieval is presented, which is extracted from the directional distance histogram of a boundary. With the descriptor, the boundary of shapes can be clearly described and the similarity between shapes can be easily computed. Experiments prove that the shape-based image retrieval method using this descriptor is efficient and very close to human´s sense
Keywords :
content-based retrieval; feature extraction; image retrieval; boundary descriptor; directional distance histogram; shape-based image retrieval; Content based retrieval; Data mining; Histograms; Image analysis; Image edge detection; Image retrieval; Image storage; Internet; Shape; Textile industry;
Conference_Titel :
Communications, 2006. APCC '06. Asia-Pacific Conference on
Conference_Location :
Busan
Print_ISBN :
1-4244-0574-2
Electronic_ISBN :
1-4244-0574-2
DOI :
10.1109/APCC.2006.255913
