Evaluating the quality of individual SIFT features

Author

Hui Su ; Wei-Hong Chuang ; Wenjun Lu ; Min Wu

Author_Institution

Dept. of Electr. & Comput. Eng., Univ. of Maryland, College Park, MD, USA

fYear

2012

fDate

Sept. 30 2012-Oct. 3 2012

Firstpage

2377

Lastpage

2380

Abstract

Scale-Invariant Feature Transform (SIFT) is one of the most popular local image features that are widely used in computer vision, image processing and image retrieval. In this paper we study the relation between the SIFT descriptor and its matching accuracy. We propose a method to quantitatively assess the quality of a SIFT feature descriptor in terms of robustness and discriminability. This would enable us to gain a better understanding of the strength and limitations of SIFT in emerging applications of SIFT-based image hash, and also to improve matching accuracy and efficiency in applications such as object search. The experimental results demonstrate the effectiveness of the proposed method.

Keywords

computer vision; feature extraction; image matching; image retrieval; SIFT features; computer vision; image hash; image processing; image retrieval; local image features; matching accuracy; scale-invariant feature transform; Accuracy; Computer vision; Estimation; Histograms; Robustness; Training; Vectors; SIFT; feature matching; feature quality; vector quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing (ICIP), 2012 19th IEEE International Conference on

Conference_Location

Orlando, FL

ISSN

1522-4880

Print_ISBN

978-1-4673-2534-9

Electronic_ISBN

1522-4880

Type

conf

DOI

10.1109/ICIP.2012.6467375

Filename

6467375