Abstract | ||
---|---|---|
This work presents a novel method performing shape matching between folded garments and corresponding unfolded templates. It incorporates both partial and global shape analysis techniques, estimating point correspondences between contours of folded garments and generic templates of unfolded garments. The established correspondences are used in the estimation of polygonal models of the folded garments. The initial matching results are also used for modifying the original templates to better fit the folded garments. The method is applied in an iterative fashion, using the fitted templates as a new input, resulting to more accurate matching as indicated by the experimental results. The produced polygon models can be used for planning the unfolding strategy during autonomous robotic manipulation. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1007/s00138-015-0671-4 | Machine Vision and Applications |
Keywords | Field | DocType |
Fold detection,Partial matching,Shape analysis,Unfolding | Computer vision,Polygon,Computer science,Artificial intelligence,Template,Shape analysis (digital geometry) | Journal |
Volume | Issue | ISSN |
26 | 4 | 0932-8092 |
Citations | PageRank | References |
0 | 0.34 | 13 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ioannis Mariolis | 1 | 33 | 4.86 |
Sotiris Malassiotis | 2 | 0 | 1.35 |