Logo des Repositoriums

Auflistung nach Autor:in "Kumar, Ajay"

1 - 3 von 3
  • Konferenzbeitrag
    Addressing biometrics security and privacy related challenges in China
    (BIOSIG 2012, 2012) Mok, Sum Yu; Kumar, Ajay
    There has been significant advancement improvement in the capabilities of biometrics and data protection technologies in last decade. The significant reduction in cost, improvements in speed and accuracy has resulted in increased deployment of such technologies in day-to-day business and public utilities. The increasing use of biometrics and data protection technologies has also raised concern on the unethical use of personal information. There are increasing number of incidents and concerns in the public over the infringement of personal privacy in China. This paper has investigated such emerging privacy related concerns in the deployment of biometrics and data protection technologies in China. This paper also includes a study on public attitudes toward such technologies and attempts to make comparison with the same in the difference with such emerging concerns in other developed countries. This paper has developed an online survey to ascertain people's understanding on the various aspects of privacy and thus willingness to tradeoff with the benefits of increased security. The online survey was conducted in February – March 2012 and revealed great deal of information from the 305 Hong Kong people. We have attempted to analyze the survey results which illustrate interesting findings on the use of CCTV, biometrics technologies, social networking, disclosure of personal information and recent (2012) privacy policy adjustments in popular websites.
  • Konferenzbeitrag
    Contactless finger knuckle identification using smartphones
    (BIOSIG 2012, 2012) Cheng, Kam Yuen; Kumar, Ajay
    This paper details the development of a smartphone based online system to automatically identify a person by using their finger knuckle image. The key objective is to exploit user-friendly biometric, with least privacy concern, to enhance security of the data in smartphone. The final product from this research is a finger knuckle authentication smartphone application, which is developed under Android operating system with environment version 2.3.3. This paper has developed some specialized algorithms for the finger knuckle detection, image preprocessing and region segmentation. Automatically detected and segmented finger knuckle images are used to encode finger knuckle pattern phase information using a pair of log-Gabor filters. Efficient implementation of various modules is achieved in C/C++ programming language, with OpenCV library, for online application. We also developed a user-friendly graphical user interface for the users to enroll and authenticate themselves. The developed system can therefore acquire finger knuckle image from the smartphone camera and automatically authenticate the genuine users. This paper has also developed a new smartphone based finger knuckle image database of 561 finger knuckle images of 187 different fingers from 109 users, in real imaging environment. In the best of our knowledge, this is the first attempt to develop a mobile phone based finger knuckle identification which has shown highly promising results in automatically identifying the users from their finger knuckle images.
  • Konferenzbeitrag
    Generating and analyzing synthetic finger vein images
    (BIOSIG 2014, 2014) Hillerström, Fieke; Kumar, Ajay; Veldhuis, Raymond N. J.
    The finger-vein biometric offers a higher degree of security, personal privacy and strong anti-spoofing capabilities than most other biometric modalities employed today. Emerging privacy concerns with the database acquisition and lack of availability of large scale finger-vein databases have posed challenges in exploring this technology for large scale applications. This paper details the first attempt to synthesize finger-vein images and presents analysis of synthesized images for the biometrics authentication. We generate a database of 50,000 fingervein images, corresponding to 5000 different subjects, with 10 different synthesized finger-vein images from each of the subject. We use tractable probability models to compare synthesized finger-vein images with the real fingervein images for their image variability. This paper also presents matching accuracy using the synthesized finger-vein database from 5000 different subjects, using 225,000 genuine and 1249,750,000 impostor matching scores, which suggests significant promises from finger-vein biometric modality for the large scale biometrics applications.