Loading...
Please wait, while we are loading the content...
Similar Documents
Semi-supervised Segmentation Fusion of Multi-spectral and Aerial Images
| Content Provider | arXiv |
|---|---|
| Author | Ozay, Mete |
| Date of Submission | 2015-02-25 |
| Abstract | A Semi-supervised Segmentation Fusion algorithm is proposed using consensus and distributed learning. The aim of Unsupervised Segmentation Fusion (USF) is to achieve a consensus among different segmentation outputs obtained from different segmentation algorithms by computing an approximate solution to the NP problem with less computational complexity. Semi-supervision is incorporated in USF using a new algorithm called Semi-supervised Segmentation Fusion (SSSF). In SSSF, side information about the co-occurrence of pixels in the same or different segments is formulated as the constraints of a convex optimization problem. The results of the experiments employed on artificial and real-world benchmark multi-spectral and aerial images show that the proposed algorithms perform better than the individual state-of-the art segmentation algorithms. |
| Related Links | https://arxiv.org/pdf/1502.04981.pdf |
| arXiv | 1502.04981 |
| Language | English |
| Access Restriction | Open |
| Subject Keyword | Computer Science - Computer Vision and Pattern Recognition Computer Science |
| Content Type | Text |
| Resource Type | Article |
| Subject | Computer Science |
