Presentation | 2013-12-12 Occlusion boundary detection based on contour shape classifiers Kazuhiko MURASAKI, Kyoko SUDO, Yukinobu TANIGUCHI, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Image segmentation is one of the basic problems for image understanding. We focus on detecting occlusion boundaries which is necessary for recognizing the region of unknown object. We propose a fast and accurate occlusion boundary detection method which simultaneously estimates the occlusion boundaries and figure-ground organization. We improve Lim's mid-level expression of boundaries so as to hold the information of boundary shapes and the relation of occlusion, and we design the estimation process to detect occlusion boundaries efficiently through the mid-level feature expression. Compared to the conventional method which uses the low-level feature such as local gradient, not only does the proposed method reduce computational time, the accuracy of figure-ground estimation is improved. The experiment show that we detect occlusion boundaries 10 times faster than conventional method, also achieves more accurately estimate the figure-ground organization. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | occlusion boundary detection / figure/ground organization / random forest / mid-level features |
Paper # | PRMU2013-77 |
Date of Issue |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2013/12/5(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Pattern Recognition and Media Understanding (PRMU) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Occlusion boundary detection based on contour shape classifiers |
Sub Title (in English) | |
Keyword(1) | occlusion boundary detection |
Keyword(2) | figure/ground organization |
Keyword(3) | random forest |
Keyword(4) | mid-level features |
1st Author's Name | Kazuhiko MURASAKI |
1st Author's Affiliation | NTT Media Intelligence Laboratories, NTT() |
2nd Author's Name | Kyoko SUDO |
2nd Author's Affiliation | NTT Media Intelligence Laboratories, NTT |
3rd Author's Name | Yukinobu TANIGUCHI |
3rd Author's Affiliation | NTT Media Intelligence Laboratories, NTT |
Date | 2013-12-12 |
Paper # | PRMU2013-77 |
Volume (vol) | vol.113 |
Number (no) | 346 |
Page | pp.pp.- |
#Pages | 5 |
Date of Issue |