Integrity Checking of Physical Currency with Pattern Matching: Coping with Few Data and the Training Sample Order

Abstract

Nowadays, technology utilization increases rapidly in our daily life. Detection of original things from fake make difficult due to misusers of this technology. Currently, world is moving toward digitalization even though physical currency is playing an important role in some of the rural areas. Counterfeit detection and maintain is a major issue. In this paper, an approach has proposed that would be useful to recognize the identity of the currencies and decreases the processing time of verification by pattern recognition techniques. This method can improves accuracy in the process of fake notes detection. In order to recognize currency features, this system uses preprocessing, color detection, segmentation, edge detection and template matching techniques. The result of this application notifies that whether the currency is original or fake to the end user. Here, we have been considering Indian currency dataset for testing the proposed system. The main strengths of this paper are performance analysis which has done by various approaches like quantitative method histogram and quality metrics comparison with test case study reports.