Boosted geometric hashing based indexing technique for finger-knuckle-print database

• This paper proposes boosted geometric hashing based indexing technique.
• This technique inserts each feature (point) exactly once into a hash table.
• The features are obtained through SIFT/SURF feature extractors.
• It reduces both computational and memory costs significantly.
• This technique tested on finger-knuckle-print (FKP).

This paper makes use of a boosted geometric hashing to propose an efficient indexing technique for finger-knuckle-print (FKP) images. Local feature extractors are used to extract features from each FKP image and each feature is inserted exactly once into the hash table which reduces memory and computational cost significantly. Features of all FKP images in the database are found to be well distributed in the hash table. It has been tested on publicly available PolyU FKP database [11] which consists of 7920 FKP images of 660 subjects. Further, the technique has been compared with a well known geometric hashing based indexing technique [6] and it is found to be better in terms of its performance.