Bi-directional Shape Correspondences (BSC)

We propose Bidirectional Shape Correspondence (BSC) as a possible improvement on the famous shape contexts (SC) framework. Our proposals derive from the observation that the SC framework enforces a one-to-one correspondence between sample points, and that this leads to two possible drawbacks. First,...

Descrición completa

Gardado en:
Detalles Bibliográficos
Autor Principal: Abdulrahman, Ibraheem
Formato: Conference paper
Idioma:Spanish / Castilian
Publicado: Abdulrahman Ibraheem 2023
Subjects:
Acceso en liña:https://ir.oauife.edu.ng/123456789/5607
Tags: Engadir etiqueta
Sen Etiquetas, Sexa o primeiro en etiquetar este rexistro!
_version_ 1810764579752378368
author Abdulrahman, Ibraheem
author_facet Abdulrahman, Ibraheem
author_sort Abdulrahman, Ibraheem
collection DSpace
description We propose Bidirectional Shape Correspondence (BSC) as a possible improvement on the famous shape contexts (SC) framework. Our proposals derive from the observation that the SC framework enforces a one-to-one correspondence between sample points, and that this leads to two possible drawbacks. First, this denies the framework the opportunity to effect advantageous many-to-many matching between points on the two shapes being compared. Second, this calls for the Hungarian algorithm which unfortunately usurps cubic time. While the dynamic-space-warping dynamic programming algorithm has provided a standard solution to the first problem above, it demands quintic time for general multi-contour shapes, and w times quadratic time for the special case of single-contour shapes, even after an heuristic search window of width w has been chosen. Therefore, in this work, we propose a simple method for computing "many-to-many" correspondences for the class of all 2-d shapes in quadratic time. Our approach is to explicitly let each point on the first shape choose a best match on the second shape, and vice versa. Along the way, we also propose the use of data-clustering techniques for dealing with the outliers problem, and, from another viewpoint, it turns out that this clustering can be seen as an autonomous, rather than pre-computed, sampling of shape boundary.
format Conference paper
id oai:ir.oauife.edu.ng:123456789-5607
institution My University
language Spanish / Castilian
publishDate 2023
publisher Abdulrahman Ibraheem
record_format dspace
spelling oai:ir.oauife.edu.ng:123456789-56072023-05-13T18:01:36Z Bi-directional Shape Correspondences (BSC) Abdulrahman, Ibraheem Bi-directional Shape Correspondences shape contexts Hungarian algorithm dynamic-space-warping dynamic programming algorithm We propose Bidirectional Shape Correspondence (BSC) as a possible improvement on the famous shape contexts (SC) framework. Our proposals derive from the observation that the SC framework enforces a one-to-one correspondence between sample points, and that this leads to two possible drawbacks. First, this denies the framework the opportunity to effect advantageous many-to-many matching between points on the two shapes being compared. Second, this calls for the Hungarian algorithm which unfortunately usurps cubic time. While the dynamic-space-warping dynamic programming algorithm has provided a standard solution to the first problem above, it demands quintic time for general multi-contour shapes, and w times quadratic time for the special case of single-contour shapes, even after an heuristic search window of width w has been chosen. Therefore, in this work, we propose a simple method for computing "many-to-many" correspondences for the class of all 2-d shapes in quadratic time. Our approach is to explicitly let each point on the first shape choose a best match on the second shape, and vice versa. Along the way, we also propose the use of data-clustering techniques for dealing with the outliers problem, and, from another viewpoint, it turns out that this clustering can be seen as an autonomous, rather than pre-computed, sampling of shape boundary. 2023-05-13T18:01:35Z 2023-05-13T18:01:35Z 2014-11 Conference paper https://ir.oauife.edu.ng/123456789/5607 es text/plain Abdulrahman Ibraheem
spellingShingle Bi-directional Shape Correspondences
shape contexts
Hungarian algorithm
dynamic-space-warping
dynamic programming algorithm
Abdulrahman, Ibraheem
Bi-directional Shape Correspondences (BSC)
title Bi-directional Shape Correspondences (BSC)
title_full Bi-directional Shape Correspondences (BSC)
title_fullStr Bi-directional Shape Correspondences (BSC)
title_full_unstemmed Bi-directional Shape Correspondences (BSC)
title_short Bi-directional Shape Correspondences (BSC)
title_sort bi directional shape correspondences bsc
topic Bi-directional Shape Correspondences
shape contexts
Hungarian algorithm
dynamic-space-warping
dynamic programming algorithm
url https://ir.oauife.edu.ng/123456789/5607
work_keys_str_mv AT abdulrahmanibraheem bidirectionalshapecorrespondencesbsc