Difference between revisions of "Publications/xu.16.pami"
From LRDE
Line 5: | Line 5: | ||
| title = Hierarchical Segmentation Using Tree-Based Shape Spaces |
| title = Hierarchical Segmentation Using Tree-Based Shape Spaces |
||
| journal = IEEE Transactions on Pattern Analysis and Machine Intelligence |
| journal = IEEE Transactions on Pattern Analysis and Machine Intelligence |
||
− | | volume = |
+ | | volume = 39 |
− | | number = |
+ | | number = 3 |
− | | pages = |
+ | | pages = 457 to 469 |
| lrdeprojects = Image |
| lrdeprojects = Image |
||
| abstract = Current trends in image segmentation are to compute a hierarchy of image segmentations from fine to coarse. A classical approach to obtain a single meaningful image partition from a given hierarchy is to cut it in an optimal way, following the seminal approach of the scale-set theory. While interesting in many cases, the resulting segmentation, being a non-horizontal cut, is limited by the structure of the hierarchy. In this paper, we propose a novel approach that acts by transforming an input hierarchy into a new saliency map. It relies on the notion of shape space: a graph representation of a set of regions extracted from the image. Each region is characterized with an attribute describing it. We weigh the boundaries of a subset of meaningful regions (local minima) in the shape space by extinction values based on the attribute. This extinction-based saliency map represents a new hierarchy of segmentations highlighting regions having some specific characteristics. Each threshold of this map represents a segmentation which is generally different from any cut of the original hierarchy. This new approach thus enlarges the set of possible partition results that can be extracted from a given hierarchy. Qualitative and quantitative illustrations demonstrate the usefulness of the proposed method. |
| abstract = Current trends in image segmentation are to compute a hierarchy of image segmentations from fine to coarse. A classical approach to obtain a single meaningful image partition from a given hierarchy is to cut it in an optimal way, following the seminal approach of the scale-set theory. While interesting in many cases, the resulting segmentation, being a non-horizontal cut, is limited by the structure of the hierarchy. In this paper, we propose a novel approach that acts by transforming an input hierarchy into a new saliency map. It relies on the notion of shape space: a graph representation of a set of regions extracted from the image. Each region is characterized with an attribute describing it. We weigh the boundaries of a subset of meaningful regions (local minima) in the shape space by extinction values based on the attribute. This extinction-based saliency map represents a new hierarchy of segmentations highlighting regions having some specific characteristics. Each threshold of this map represents a segmentation which is generally different from any cut of the original hierarchy. This new approach thus enlarges the set of possible partition results that can be extracted from a given hierarchy. Qualitative and quantitative illustrations demonstrate the usefulness of the proposed method. |
||
− | | note = To appear |
||
| lrdepaper = http://www.lrde.epita.fr/dload/papers/xu.16.pami.pdf |
| lrdepaper = http://www.lrde.epita.fr/dload/papers/xu.16.pami.pdf |
||
| lrdekeywords = Image |
| lrdekeywords = Image |
||
Line 23: | Line 22: | ||
journal = <nowiki>{</nowiki>IEEE Transactions on Pattern Analysis and Machine |
journal = <nowiki>{</nowiki>IEEE Transactions on Pattern Analysis and Machine |
||
Intelligence<nowiki>}</nowiki>, |
Intelligence<nowiki>}</nowiki>, |
||
− | year = <nowiki>{</nowiki> |
+ | year = <nowiki>{</nowiki>2017<nowiki>}</nowiki>, |
− | volume = <nowiki>{</nowiki> |
+ | volume = <nowiki>{</nowiki>39<nowiki>}</nowiki>, |
− | number = <nowiki>{</nowiki> |
+ | number = <nowiki>{</nowiki>3<nowiki>}</nowiki>, |
− | pages = <nowiki>{</nowiki> |
+ | pages = <nowiki>{</nowiki>457--469<nowiki>}</nowiki>, |
+ | month = apr, |
||
abstract = <nowiki>{</nowiki>Current trends in image segmentation are to compute a |
abstract = <nowiki>{</nowiki>Current trends in image segmentation are to compute a |
||
hierarchy of image segmentations from fine to coarse. A |
hierarchy of image segmentations from fine to coarse. A |
||
Line 50: | Line 50: | ||
from a given hierarchy. Qualitative and quantitative |
from a given hierarchy. Qualitative and quantitative |
||
illustrations demonstrate the usefulness of the proposed |
illustrations demonstrate the usefulness of the proposed |
||
− | method.<nowiki>}</nowiki> |
+ | method.<nowiki>}</nowiki> |
− | note = <nowiki>{</nowiki>To appear<nowiki>}</nowiki> |
||
<nowiki>}</nowiki> |
<nowiki>}</nowiki> |
||
Revision as of 15:08, 9 February 2017
- Authors
- Yongchao Xu, Edwin Carlinet, Thierry Géraud, Laurent Najman
- Journal
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Type
- article
- Projects
- Image"Image" is not in the list (Vaucanson, Spot, URBI, Olena, APMC, Tiger, Climb, Speaker ID, Transformers, Bison, ...) of allowed values for the "Related project" property.
- Keywords
- Image
- Date
- 2016-04-11
Abstract
Current trends in image segmentation are to compute a hierarchy of image segmentations from fine to coarse. A classical approach to obtain a single meaningful image partition from a given hierarchy is to cut it in an optimal way, following the seminal approach of the scale-set theory. While interesting in many cases, the resulting segmentation, being a non-horizontal cut, is limited by the structure of the hierarchy. In this paper, we propose a novel approach that acts by transforming an input hierarchy into a new saliency map. It relies on the notion of shape space: a graph representation of a set of regions extracted from the image. Each region is characterized with an attribute describing it. We weigh the boundaries of a subset of meaningful regions (local minima) in the shape space by extinction values based on the attribute. This extinction-based saliency map represents a new hierarchy of segmentations highlighting regions having some specific characteristics. Each threshold of this map represents a segmentation which is generally different from any cut of the original hierarchy. This new approach thus enlarges the set of possible partition results that can be extracted from a given hierarchy. Qualitative and quantitative illustrations demonstrate the usefulness of the proposed method.
Documents
Bibtex (lrde.bib)
@Article{ xu.16.pami, author = {Yongchao Xu and Edwin Carlinet and Thierry G\'eraud and Laurent Najman}, title = {Hierarchical Segmentation Using Tree-Based Shape Spaces}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {2017}, volume = {39}, number = {3}, pages = {457--469}, month = apr, abstract = {Current trends in image segmentation are to compute a hierarchy of image segmentations from fine to coarse. A classical approach to obtain a single meaningful image partition from a given hierarchy is to cut it in an optimal way, following the seminal approach of the scale-set theory. While interesting in many cases, the resulting segmentation, being a non-horizontal cut, is limited by the structure of the hierarchy. In this paper, we propose a novel approach that acts by transforming an input hierarchy into a new saliency map. It relies on the notion of shape space: a graph representation of a set of regions extracted from the image. Each region is characterized with an attribute describing it. We weigh the boundaries of a subset of meaningful regions (local minima) in the shape space by extinction values based on the attribute. This extinction-based saliency map represents a new hierarchy of segmentations highlighting regions having some specific characteristics. Each threshold of this map represents a segmentation which is generally different from any cut of the original hierarchy. This new approach thus enlarges the set of possible partition results that can be extracted from a given hierarchy. Qualitative and quantitative illustrations demonstrate the usefulness of the proposed method.} }