Fast approximation of Parzen density estimation for fingerprint alignment

作者：Park Heemin^*

来源：IEEJ Transactions on Electrical and Electronic Engineering, 2012, 7: S182-S183.

DOI：10.1002/tee.21822

• fingerprint matching
• alignment
• minutiae
• Parzen density estimation

Previous fingerprint alignment approaches may produce biased results because of outliers caused by fingerprint noise. To account for noise, major trends from transformations can be extracted using Parzen density estimation. However, this extraction is computationally intense. Here, we propose a fast approximation algorithm of the Parzen density estimation for global fingerprint alignment. Experimental results show that the proposed algorithm's performance is as good as that of Parzen density estimation and it has a much shorter execution time.

• 出版日期2012-12