Edinburgh Napier University

  Objective:
Hand vein images have become important biometric signs used for identification systems. Also, dorsal hand vein images have noteworthy advantages in terms of reliability and contactless procedure. Surgically changing the vascular pattern under the skin is extremely difficult. Therefore, the use of such patterns in identity recognition applications is increasing day by day. In this context, many studies have been carried out in essential areas such as image acquisition for different textures, image preprocessing, data security, image feature extraction and recognition.

Methods:
In this article, a new dorsal hand vein identification application has been carried out with the chaos-based security mechanism to protect personal data and the Improved SURF method to reduce the error matches of traditional SURF method in the current application. There are two improvements in the presented system. The first one finds the features representing only the vein segments, not the remaining tissue or hairly parts. The second one eliminates the mismatching points to increase the accuracy.

Results:
Both encryption and data hiding techniques are preferred together to protect person data. In addition, dorsal hand vein images taken with a camera for matching and recorded in the database of the people have been improved with the newly developed I-SURF feature extraction method. Thereby, with the developed application, in addition to keeping the personal identity information securely in the database with chaos-based methods, higher accuracy rate has been achieved in matching the new image taken with the camera.