BROWSE
Detail
Conference Paper
Magnetic Resonance Image Retrieval Based on Contour to Centroid Triangulation with Shape Feature Similarity
Cited 1 time in 
Share 
Share
- Authors
- Byung K. Jung, Seong H. Son, Jeong K. Pack
- Issue Date
- 2013-10
- Citation
- Research in Adaptive and Convergent Systems (RACS) 2013, pp.182-186
- Language
- English
- Type
- Conference Paper
- Abstract
- In this paper, we present an image retrieval method based on contour to centroid triangulation with shape feature similarity. We assume test images and database images used in this paper are all single objects that are segmented by known algorithms such as SVM and K-means algorithms. From these classified binary images, we propose novel Shape based image retrieval method integrating sectored characteristic points to the Contour to Centroid Triangulation (CTCT) method using Unique Representation Grid (URG) as shape feature that can perform as the filtering process. The experimental result shows proposed method has improved conventional CTCT in retrieving medical object image compared to conventional CTCT method with 79 percent match rate while CTCT showed 33 percent match. © 2013 ACM.
- KSP Keywords
- Experimental Result, Image retrieval, Magnetic resonance(MR), Magnetic resonance images, Novel shape, Object image, Single objects, binary image, characteristic points, feature similarity, filtering process