Reducing Computational Cost of Pair-Polar Coordinate-based Cancelable Fingerprint Template Matching

Abstract

Since a long time ago, biometrics has become a popular authentication method to verify users easily. Despite this advantage, the user’s biometric data must be stored securely because those data cannot be replaced once it has been compromised. It is because humans cannot alter parts of his or her body. Many methods can be used to secure biometric data. One of them is the cancelable fingerprint template. However, its matching computation process is quite expensive, as many minutia points need to compare. In this paper, we optimize the number of minutia points in its selection process using a new threshold. It is to reduce the computational cost of pair-polar coordinate-based cancelable fingerprint template’s matching. The experiment result shows that the proposed method achieves five times better performance than the previous method.