Addressing Sensor Interoperability Problem using Fingerprint Segmentation

Varnika Varnika, Chander Kant


The advances in sensor technology allow us to acquire fingerprint data of a person through variety of fingerprint sensors. With the advancing technology, it is common to replace older designs with newer ones. So, it is possible that the sensors used at the time of fingerprint enrollment and identification may be different. Interoperability refers to the system’s ability to work with various set of devices. There should be interoperability between sensors otherwise the performance of recognition system will be affected. In this paper fingerprint sensor interoperability problem is addressed using fingerprint segmentation. Segmentation is one of the most important phases in recognition of any biometric modality. It is the process of separating the foreground (which contains the information) and the background (which is basically the noise in the image). The devices available today generally work accurately if the image is acquired from a certain sensor. The problem occurs if the acquiring device is changed and the background of the image can wrongly be labeled as foreground in the segmentation process. This problem is discussed and a solution to overcome this problem is presented in this paper.

Keywords: Background Component, Biometric System, Fingerprint Sensors, Foreground Component, Segmentation, Sensor Interoperability.

Full Text:




  • There are currently no refbacks.

Copyright (c) 2016 International Journal of Advanced Research in Computer Science