TY - GEN
T1 - Near-duplicate detection using a new framework of constructing accurate affine invariant regions
AU - Tian, Li
AU - Kamata, Sei Ichiro
PY - 2007/1/1
Y1 - 2007/1/1
N2 - In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions and use it for near-duplicate detection problem. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations are extracted from seed points by a new method named the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours of TSGRs to obtain ellipse regions as the final invariant regions. At last, SIFT-PCA descriptors are computed on the obtained regions. In the experiment, our framework is evaluated by retrieving near-duplicate in an image database containing 1000 images. It gives a satisfying result of 96.8% precision at 100% recall.
AB - In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions and use it for near-duplicate detection problem. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations are extracted from seed points by a new method named the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours of TSGRs to obtain ellipse regions as the final invariant regions. At last, SIFT-PCA descriptors are computed on the obtained regions. In the experiment, our framework is evaluated by retrieving near-duplicate in an image database containing 1000 images. It gives a satisfying result of 96.8% precision at 100% recall.
KW - Ellipse fitting
KW - Image matching
KW - Invariant region
KW - Near-duplicate detection
KW - Thresholding seeded growing regions
UR - http://www.scopus.com/inward/record.url?scp=38349030529&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=38349030529&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-76414-4_7
DO - 10.1007/978-3-540-76414-4_7
M3 - Conference contribution
AN - SCOPUS:38349030529
SN - 9783540764137
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 61
EP - 72
BT - Advances in Visual Information Systems - 9th International Conference, VISUAL 2007, Revised Selected Papers
PB - Springer Verlag
T2 - 9th International Conference on Visual Information Systems, VISUAL 2007
Y2 - 28 June 2007 through 29 June 2007
ER -