Wavelet compression of fingerprints and recognition using moment invariants

  • Authors:

  • Tsun-Kuo Lin

  • Affiliations:

  • Department of Information Management, Shih Chien University Kaohsiung Campus, Neimen Shiang, Kaohsiung County, Taiwan

  • Venue:
  • AIA'06 Proceedings of the 24th IASTED international conference on Artificial intelligence and applications
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Identification and efficient data compression of fingerprint images are both essential for automatic fingerprint recognition system due to the increasing demand on storage space and the slow exchange of the images between agencies. In this work, moment invariants combined with wavelet filters for fingerprint recognition are proposed to reduce the transmission cost while preserving the person's identity. It is found that high data compression and high recognition rates can be achieved using this method. In the experimental ranges, the most suitable wavelet filter for the fingerprint recognition is conducted. The evaluation of wavelet compression is presented using root-mean-square signal-to-noise ratio, which can be modeled as the relation of first-order exponential decay with compression scale index.