Joint Utilization of Appearance and Geometry for Determining Correspondences

Date

2009-09-16

Author

Soysal, Medeni
Alatan, Abdullah Aydın
Karadeniz, Talha

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

198
views

0
downloads

A novel approach, which is based on combining the competence of interest point detectors to capture primitives and the capability of geometric constraints to discriminate between spatial configurations of these primitives is presented. In the proposed approach, the geometric constraints are enforced by means of barycentric coordinates, a mathematical tool that has been utilized in the relevant literature as a neighborhood constraint. In the context of our research, however, these coordinates are utilized as a geometric description of a group of interest points. Using this description, local appearance descriptor based potential matches are intended to be filtered and evaluated according to their geometric consistency. This method is applicable to one-to-one image matching in its current form, and to classification tasks after extensions for appearance generalization.

Subject Keywords

Interest point, Local region description, Geometrically consistent groups, Barycentric coordinates

URI

https://hdl.handle.net/11511/37827

DOI

https://doi.org/10.1109/iscis.2009.5291918

Collections

Department of Electrical and Electronics Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Analysis and design of conformal frequency selective surfaces / Dalkılıç, Akın; Alatan, Lale; Department of Electrical and Electronics Engineering (2014) An efficient analysis and design approach for conformal frequency selective surface (FSS) structures is developed. The design methodology involves the analysis of both the planar and curved FSS structures. First, planar unit cell analysis of conformal FSS models are accomplished for normal and oblique incidence cases. To prove conformal applicability of planar designs, a semi-finite analysis method is utilized. This method is based on solution of a singly periodic curved FSS structure of semi-cylinder shape...
Analysis Window Length Selection For Linear Signal Models Yazar, Alper; Candan, Çağatay (2015-05-19) A method is presented for the selection of analysis window length, or the number of input samples, for linear signal modeling without compromising the model assumptions. It is assumed that the signal of interest lies in a known linear space and noisy samples of the signal is provided. The goal is to use as many signal samples as possible to mitigate the effect of noise without violating the assumptions on the model. An application example is provided to illustrate the suggested method.
Analysis of Face Recognition Algorithms for Online and Automatic Annotation of Personal Videos Yılmaztürk, Mehmet; Ulusoy Parnas, İlkay; Çiçekli, Fehime Nihan (Springer, Dordrecht; 2010-05-08) Different from previous automatic but offline annotation systems, this paper studies automatic and online face annotation for personal videos/episodes of TV series considering Nearest Neighbourhood, LDA and SVM classification with Local Binary Patterns, Discrete Cosine Transform and Histogram of Oriented Gradients feature extraction methods in terms of their recognition accuracies and execution times. The best performing feature extraction method and the classifier pair is found out to be SVM classification...
Numerical Solution of Multi-scale Electromagnetic Boundary Value Problems by Utilizing Transformation-Based Metamaterials Ozgun, Ozlem; Kuzuoğlu, Mustafa (2011-06-23) We present numerical solution techniques for efficiently handling multi-scale electromagnetic boundary value problems having fine geometrical details or features, by utilizing spatial coordinate transformations. The principle idea is to modify the computational domain of the finite methods (such as the finite element or finite difference methods) by suitably placing anisotropic metamaterial structures whose material parameters are obtained by coordinate transformations, and hence, to devise easier and effic...
Electromagnetic target recognition with the fused MUSIC spectrum matrix method: Applications and performance analysis for incomplete frequency data Secmen, Mustafa; Ekmekci, Evren; Sayan, Gönül (2007-01-01) The aim of this paper is to apply an electromagnetic target recognition method, which is based on the use of fused MUSIC spectrum matrices, to the case of incomplete frequency domain data. The aforementioned method was suggested recently and succesfully applied to both canonical and complicated targets in the presence of complete frequency domain data [1]. However, most of the real world applications involve the use of severely incomplete frequency data, especially missing low frequency information. In this...

Citation Formats

M. Soysal, A. A. Alatan, and T. Karadeniz, “Joint Utilization of Appearance and Geometry for Determining Correspondences,” 2009, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/37827.