Multidimensional binary search trees used for associative searching
Communications of the ACM
The K-D-B-tree: a search structure for large multidimensional dynamic indexes
SIGMOD '81 Proceedings of the 1981 ACM SIGMOD international conference on Management of data
Distinctive Image Features from Scale-Invariant Keypoints
International Journal of Computer Vision
Speeded-Up Robust Features (SURF)
Computer Vision and Image Understanding
Robust iris indexing scheme using geometric hashing of SIFT keypoints
Journal of Network and Computer Applications
Iris-Biometric Hash Generation for Biometric Database Indexing
ICPR '10 Proceedings of the 2010 20th International Conference on Pattern Recognition
An iris retrieval technique based on color and texture
Proceedings of the Seventh Indian Conference on Computer Vision, Graphics and Image Processing
Real-time iris segmentation based on image morphology
Proceedings of the 2011 International Conference on Communication, Computing & Security
Indexing biometric databases using pyramid technique
AVBPA'05 Proceedings of the 5th international conference on Audio- and Video-Based Biometric Person Authentication
A Fast Search Algorithm for a Large Fuzzy Database
IEEE Transactions on Information Forensics and Security
IEEE Transactions on Circuits and Systems for Video Technology
Hi-index | 0.00 |
In this paper, an indexing approach is proposed for clustered SIFT keypoints using k-d-b tree. K-d-b tree combines the multidimensional capability of k-d tree and balancing efficiency of B tree. During indexing phase, each cluster center is used to traverse the region pages of k-d-b tree to reach an appropriate point page for insertion. For m cluster centers, m such trees are constructed. Insertion of a node into k-d-b tree is dynamic that generates balanced data structure and incorporates deduplication check as well. For retrieval, range search approach is used which finds the intersection of probe cluster center with each region page being traversed. The iris identifiers on the point page referenced by probe iris image are retrieved. Results are obtained on publicly available BATH and CASIA Iris Image Database Version 3.0. Empirically it is found that k-d-b tree is preferred over state-of-the-art biometric database indexing approaches.