Title :
An Efficient and Robust Technique for Region Based Shape Representation and Retrieval
Author :
Zhang, Dengsheng ; Lim, Melissa Chen Yi
Author_Institution :
Monash Univ., Churchill
Abstract :
Most region based shape image retrieval techniques in literature are complex to implement. As the result of their complex computations, the retrieval performance of these techniques is not robust. In this paper, we propose a new approach to use contour based method for region based shape representation and retrieval. Specifically, we compute a one dimensional signature function from a two dimensional region shape and use it for region shape representation and retrieval. The proposed region shape feature is much more efficient to compute than conventional region shape techniques. The retrieval performance of the technique has been evaluated using a standard region shape database. Experimental results show that the proposed technique is efficient and robust, and outperforms the widely used shape methods in literature.
Keywords :
feature extraction; image retrieval; contour based method; one dimensional signature function; region based shape representation; region shape feature; shape image retrieval; two dimensional region shape; Australia; Image databases; Image retrieval; Image sampling; Information retrieval; Information technology; MPEG 7 Standard; Robustness; Shape measurement; Sparse matrices; CBIR; Image retrieval; PRS.; shape; signature;
Conference_Titel :
Computer and Information Science, 2007. ICIS 2007. 6th IEEE/ACIS International Conference on
Conference_Location :
Melbourne, Qld.
Print_ISBN :
0-7695-2841-4
DOI :
10.1109/ICIS.2007.55