Abstract | ||
---|---|---|
In this paper we propose a three-stage algorithm to implement effective segmentation of an object's counterparts when illumination variance is present. A color constancy algorithm and color-based invariant parameters insensitive to a large set of different illumination conditions are used. Then reflectance images are considered after discarding the shadowing information present in the images. A color-based segmentation algorithm using Graph Cuts is applied to the reflectance images. Improvements in the segmentation are obtained after using these illumination invariants. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1007/978-3-319-11758-4_20 | IMAGE ANALYSIS AND RECOGNITION, ICIAR 2014, PT I |
Keywords | Field | DocType |
Color-invariants, Graph cuts, Reflectance, Intrinsic images | Cut,Computer vision,Color constancy,Scale-space segmentation,Pattern recognition,Segmentation,Computer science,Segmentation-based object categorization,Photometry (optics),Image segmentation,Invariant (mathematics),Artificial intelligence | Conference |
Volume | ISSN | Citations |
8814 | 0302-9743 | 1 |
PageRank | References | Authors |
0.34 | 7 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
José Antonio Pérez-Carrasco | 1 | 21 | 5.49 |
Begoña Acha-Piñero | 2 | 4 | 0.76 |
Carmen Serrano-Gotarredona | 3 | 1 | 0.34 |
Theo Gevers | 4 | 2973 | 214.46 |