CNN based Finger Region Segmentation for Finger Vein Recognition
This is "The Multimedia Signal Processing and Security Lab", short WaveLab, website. We are a research group at the Artificial Intelligence and Human Interfaces (AIHI) Department of the University of Salzburg led by Andreas Uhl. Our research is focused on Visual Data Processing and associated security questions. Most of our work is currently concentrated on Biometrics, Media Forensics and Media Security, Medical Image and Video Analysis, and application oriented fundamental research in digital humanities, individualised aquaculture and sustainable wood industry.
Finger region segmentation is an important step in a biometric finger vein recognition toolchain. Its aim is to separate the finger region from background and all other objects of the image. So far, finger region extraction for finger vein recognition systems has mainly used classical image processing based systems. In this work three state-of-the art convolutional neural network (CNN) based architectures for segmentation, namely Mask R-CNN, CCNet and HRNet, are evaluated. A major advantage of the presented CNN-based approach compared to classic image processing approaches is that the images neither have to be preprocessed nor any parameter have to be optimized. All that is required is a sufficient number of already segmented finger vein images for training.
Coming soon ...
The Hong Kong Polytechnic University Finger Image Database
The Hong Kong Polytechnic University Finger Image Database (HKPU) Database consists of simultaneously acquired finger vein and finger surface texture images. It contains finger vein images from 156 volunteers, four fingers each (left and right index and middle finger). The data was captured in two different sessions, capturing six samples per finger in each session. Further information regarding the data set can be found by the following link:
HKPU Download Page (external link)
PLUSVein-Contactless Finger and Hand Vein Data Sett
The PLUSVein-CL data set consists of 3 subsets: a palmar finger vein one, acquired using the light transmission illuminator and two hand vein ones, one acquired with the 850 nm reflected light illuminator and the other one with the 950 nm reflected light illuminator. Currently it contains 42 subjects. 6 fingers (left and right index, middle and ring finger) and two hands (left and right hand) per subject and 5 images per finger/hand in 1 session were captured for each subsets.. Further information regarding the data set can be found by the following link:
PROTECT MultiModal Data Set
The PROTECT multimodal data set (PMMDB) Database consists of simultaneously acquired finger vein and finger surface texture images. The PMMDB includes different biometric modalities, namely iris, face (visual light, NIR, 3D and thermal), periocular, anthropometrics and hand- and finger veins of 69 different subjects. It was acquired in two data acquisition events with one year between the two sessions. Further information regarding the data set can be found by the following link:
PMMDB Download Page (external link)
University of Twente Finger Vascular Pattern Database
The University of Twente Finger Vascular Pattern (UTFVP) Database is a publically available finger vein data set. It contains six fingers (ring, middle and index finger from both hands) from 60 volunteers acquired in two sessions. Further information regarding the data set can be found by the following link:
UTFVP Download Page (external link)
DOWNLOAD: Ground truth for HKPU, PMMDB, UTFVP
The download files are available upon request. Please fill out this form to request a download link for the manually segmented finger region masks: