Pose-Specific 3D Fingerprint Unfolding

TL;DR

The paper tackles the challenge of making 3D fingerprint data compatible with traditional 2D fingerprint systems by addressing pose-induced deformation. It introduces a pose-specific unfolding pipeline that first unfolds a 3D fingerprint in its original pose, estimates the 3D pose of the query to align with the flat fingerprint, and unfolds again in that pose before standard 2D registration. Experiments on a newly collected dataset show that pose-specific unfolding reduces distortion and yields higher genuine matching scores compared to general unfolding, though the approach is relatively slow and requires per-query processing. This work enhances cross-domain fingerprint matching by aligning 3D data to the 2D recognition framework, enabling more reliable enrollment and verification across sensor modalities.

Abstract

In order to make 3D fingerprints compatible with traditional 2D flat fingerprints, a common practice is to unfold the 3D fingerprint into a 2D rolled fingerprint, which is then matched with the flat fingerprints by traditional 2D fingerprint recognition algorithms. The problem with this method is that there may be large elastic deformation between the unfolded rolled fingerprint and flat fingerprint, which affects the recognition rate. In this paper, we propose a pose-specific 3D fingerprint unfolding algorithm to unfold the 3D fingerprint using the same pose as the flat fingerprint. Our experiments show that the proposed unfolding algorithm improves the compatibility between 3D fingerprint and flat fingerprint and thus leads to higher genuine matching scores.

Figures (7)

• Figure 1: Flowchart of the proposed pose-specific 3D fingerprint unfolding.
• Figure 2: Schematic diagram of the visualization method.
• Figure 3: Visualization results of 3D point clouds (left) and real flat fingerprints (right) in each subfigure. Image quality is similar, while the former has a larger area. There is perspective distortion in the visualization result which will be removed by unfolding step.
• Figure 4: Visualization results of point clouds of two fingers of different poses before (left) and after (right) unfolding.
• Figure 5: Illustration of 3D pose estimation. (a) A real flat fingerprint. (b) Corresponding finger point cloud in the original pose. (c) Corresponding finger point cloud in the same pose as the flat fingerprint.