Visual saliency based on selective integration of feature maps in frequency domain
- Title
- Visual saliency based on selective integration of feature maps in frequency domain
- Author
- 문영식
- Keywords
- Visualization; Automatic method; Multiple objects; Spectral entropy; Cluttered backgrounds; Feature map; Visual saliency; Frequency domains; Consumer electronics; Frequency domain analysis; Natural images
- Issue Date
- 2013-01
- Citation
- Digest of Technical Papers - IEEE International Conference on Consumer Electronics, article no. 6486787, Page. 43-44
- Abstract
- In this paper, an automatic method for extracting visual saliency based on selective integration of feature maps in frequency domain is proposed. Feature maps are calculated by measuring the Bayes spectral entropy. In order to extract visual saliency effectively, feature maps are first generated from three images separated into Y, Cb, Cr channels, respectively. Then, by selectively integrating feature maps, visual saliency is finally extracted. Experimental results have shown that the proposed method obtains good performance of visual saliency under various environments containing multiple objects and cluttered backgrounds in natural images. © 2013 IEEE.
- URI
- https://ieeexplore.ieee.org/document/6486787/https://repository.hanyang.ac.kr/handle/20.500.11754/185567
- ISSN
- 0747-668X
- DOI
- 10.1109/ICCE.2013.6486787
- Appears in Collections:
- COLLEGE OF COMPUTING[E](소프트웨어융합대학) > COMPUTER SCIENCE(소프트웨어학부) > Articles
- Files in This Item:
There are no files associated with this item.
- Export
- RIS (EndNote)
- XLS (Excel)
- XML