Introduction

Abstract

The rapid emergence of several terror groups and organizations in the past few decades, human security and protection has now been assumed to be under severe threat. In present scenario, human identification and recognition systems can be seen as a big potential preventive measures. Apart from that, government agencies have been trying to initiate a robust linkage between several financial inclusion (such as banking, insurance)/ public welfare schemes (such as NEREGA and MENREGA in Indian scenario) and biometric based personal authentication mechanism. This is to ensure corruption free realization and management of such policies to the entire society. There are several biometric traits like face, iris, fingerprint, palmprint etc., but fingerprint has been widely used and accepted modality. It has been observed that most of the time fingerprint quality of laborers and cultivators is very poor due to continuous rough hand usage. This will seriously affect any such initiation adversely, especially in India and other Asian countries, that are primarily rural. But on the other hand, finger knuckle image quality has been surprisingly observed to be very good, as no one has use them for any work. In this paper, we are proposing a novel finger-knuckle-print based biometric system to check the escapes that are present in transfer of payments through the various levels of bureaucratic financial inclusion projects. Initially, ROI has been extracted, enhanced and transformed using proposed BOP,SOP and IRT based locally adapted procedures. Then, CNN based DeepMatching technique has been used to match multiple features of two FKP images. Finally, score level fusion rule has been done revealing improvement in the system performance..

Motivation

In several Asian countries like India more than 60% of the population reside in rural areas. It has been observed that their quality of fingerprint is not very good. The laborers, cultivators do substantial work and use their hands very roughly. This causes plenty of damage to their fingerprint permanently. In such scenario, the quality of FKP is unaffected because they can not be used for any other purpose and hence less prone to injuries.

Objective

Given the original FKP data of human finger, Segment it automatically into ROI having similar features which are ​anatomically meaningful​​. Make a software, which will take FKP data of finger as input and gives ​similar feature FKP image and some useful information according to given application like-for Image Transformation, it will give us shortest path for ROI images to enhance the FKP patterns.

Keywords:

  • ROI Enancement
  • Image Transformation
  • DeepMatching