Loading...
Please wait, while we are loading the content...
A Two Stage Approach to Detect Salient Objects in Noisy Images
| Content Provider | Semantic Scholar |
|---|---|
| Author | Kumar, Nitin Singh, Maheep |
| Copyright Year | 2016 |
| Abstract | Salient Object Detection (SOD) has become an active research area in past two decades due to its critical applications in image segmentation, image compression, object discovery and video summarization etc. In literature, several methods have been proposed to detect salient objects in digital images. The main objective of salient object detection methods is to extract visually dissimilar areas in digital images which are free from any artifact e.g. noise. In this paper, we have proposed a novel two stage approach to detect salient object in noisy images. In the first stage, we have applied two-dimensional wiener filter to reduce noise affect in the image while in the second stage, features are extracted using Liu et al., method. The effectiveness of the proposed approach is demonstrated on two publicly available datasets viz. ASD and MSRA5K. Experimental Results demonstrate that the proposed approach outperforms several other methods in terms of Precision, Recall, F-measure and Area-Under the Curve. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://www.ijipbangalore.org/abstracts_10(2)/p1.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
