BROWSE
Detail
Conference Paper
Similarity Measurement with Mesh Distance Fourier Transform in 2D Binary Image
Cited 2 time in 
Share 
Share
- Authors
- Ravi Kasaudhan, Tae K. Heo, Soon Ik Jeon, Seong Ho Son
- Issue Date
- 2015-10
- Citation
- Research in Adaptive and Convergent Systems (RACS) 2015, pp.183-148
- Language
- English
- Type
- Conference Paper
- Abstract
- Shape based feature is a widely used method in Content based Image Retrieval (CBIR) for similarity measurements because contours of an image provide relevant information for similarity. In this paper, we propose a novel shape feature named Mesh Distance Fourier Descriptor (MDFD) which takes into account the contour information of each of the boundary points with respect to other contour points in the images such that the relationship of one boundary point is evaluated with respect to all other boundary points in 2D space. In this paper we have used binary images which are classified into single objects using known classification methods such as K-means and SVM algorithms. The proposed method has been compared with Sectorized Object Matching (SOM) and the result shows that the proposed algorithm outperforms SOM in terms of matching of similar images.
- KSP Keywords
- 2D space, Boundary points, Classification method, Content based image retrieval (cbir), Fourier Descriptor, Fourier Transform, K-Means, Mesh distance, Novel shape, Similarity Measurement, Single objects