Decision-level fusion in fingerprint verification

By Salil Prabhakar, Anil K. Jain. Pattern Recognition, 2001

438 citations recorded by the source corpus.

A scheme is proposed for classifier combination at decision level which stresses the importance of classifier selection during combination. The proposed scheme is optimal (in the Neyman-Pearson sense) when sufficient data are available to obtain reasonable estimates of the join densities of classifier outputs. Four different fingerprint matching algorithms are combined using the proposed scheme to improve the accuracy of a fingerprint verification system. Experiments conducted on a large fingerprint database (∼ 2,700 fingerprints) confirm the effectiveness of the proposed integration scheme. An overall matching performance increase of ∼ 3% is achieved. We further show that a combination of multiple impressions or multiple fingers improves the verification performance by more than 4% and 5%, respectively. Analysis of the results provide some insight into the various decision-level classifier combination strategies.

Subfields: Computer Vision and Pattern Recognition, Safety Research, Signal Processing

View Paper: https://www.semanticscholar.org/paper/0262c1875b8fb04c43de92cff13fe9ca8a0ad483

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Safety Research

In Signal Processing

Citations in the corpus

Guide to Biometrics (2003, methodology): “For example, Prabhakar and Jain [172] estimate the conditional densities using a non-parametric estimation method (e.”
An introduction to biometric recognition (2004, background): “Jain and Ross improved the performance of a multimodal biometric system by learning user-specific parameters [30].”
Multimodal biometrics: An overview (2004, background): “Figure 2: Scenarios in a multimodal biometric system [6].”
Likelihood Ratio-Based Biometric Score Fusion (2008, methodology): “This approach is general and is able to minimize the FRR at a specified FAR.”
Biometrics: a tool for information security (2006, methodology): “the joint generalized densities obtained by using copula fusion rule result in improved matching performance as opposed to product or marginal density estimates (see [40] and [66]).”
Efficient Privacy-Preserving Biometric Identification (2011, background): “In the last decade, many fingerprint-recognition techniques have been developed that combine various features of the fingerprint [2, 18, 19, 21].”
Robustness of multi-modal biometric verification systems under realistic spoofing attacks (2011, methodology): “To this aim, we fit the available genuine and impostor matching scores with a parametric distribution [18, 19, 12].”
Fingerprint recognition by combining global structure and local cues (2006, methodology): “A more general method is to directly estimate the two-dimensional (2-D) distributions such as the one proposed in [27] using 2-D Parzen Window.”
Score normalization in multimodal biometric systems (2005, background): “Jain and Ross [22] have proposed the use of user-specific weights for computing the weighted sum of scores from the different modalities.”
Information fusion in biometrics (2003, methodology): “A brief description of the three biometric indicators used in our experiments is given below.”