FingerNet: An unified deep network for fingerprint minutiae extraction

Abstract

Minutiae extraction is of critical importance in automated fingerprint recognition.Previous works on rolled/slap fingerprints failed on latent fingerprints due to noisy ridge patterns and complex background noises. In this paper, we propose a new way to design deep convolutional network combining domain knowledge and the representation ability of deep learning. In terms of orientation estimation, segmentation, enhancement and minutiae extraction, several typical traditional methods performed well on rolled/slap fingerprints are transformed into convolutional manners and integrated as an unified plain network. We demonstrate that this pipeline is equivalent to a shallow network with fixed weights. The network is then expanded to enhance its representation ability and the weights are released to learn complex background variance from data, while preserving end-to-end differentiability. Experimental results on NIST SD27 latent database and FVC 2004 slap database demonstrate that the proposed algorithm outperforms the state-of-the-art minutiae extraction algorithms. Code is made publicly available at: https://github.com/felixTY/FingerNet.

Robust Minutiae Extractor: Integrating Deep Networks and Fingerprint Domain Knowledge

“…Unlike existing works using plain convolutional neural network [21,22] or sliding window [18,3] to process each patch with fixed size and stride, we use a deeper residual learning based network with more pooling layers to scale down the region patch. Specifically, we get the output after the 2 nd , 3 rd , and 4 th pooling layer to feed to an ASPP network [2] with corresponding rates for multiscale segmentation.…”

“…Using non-maximum suppression to reduce the number of candidates is common in object detection [7,17]. Some of the candidate regions are deleted to get a reliable minutiae score map by setting a hard threshold [3] or using heuristics [21,22]. However, a hard threshold can also suppress valid minutiae locations.…”

“…It also suffers from time-consuming sliding windows as in [13], and separate modules for minutiae location and orientation estimates. Tang et al [22] proposed FingerNet that maps traditional minutiae extraction pipeline including orientation estimation, segmentation, enhancement, and extraction to a network with fixed weights. Although this approach is promising because it combines domain knowledge and deep network, it still uses plain 3 network architecture and hard threshold in non-maximum suppression 4 .…”

FPD-M-net: Fingerprint Image Denoising and Inpainting Using M-net Based Convolutional Neural Networks

“…Cao et al [1] pose latent orientation estimation as a patch classification problem using CNN. Tang et al [22] proposed a FingerNet, based on deep convolutional network. It uses domain knowledge for fingerprint minutiae extraction in noisy ridge patterns and complex background.…”

U-Finger: Multi-Scale Dilated Convolutional Network for Fingerprint Image Denoising and Inpainting

“…Convolutional neural networks (CNNs) [20] have brought unprecedented success in many computer vision tasks, including some recent works addressing fingerprint extraction and analysis [1,2]. On the other hand, fingerprint restoration and enhancement have been traditionally studied using classical example-based and regression methods [3,4,5,8] [6,10,11,17,18], the recent ECCV 2018 ChaLearn competition 3 has started to motivate researchers to develop deep learning algorithms that can restore fingerprint images that contain artifacts such as noise, scratches [7,9], etc.…”