Sri Lanka Currency Note Recognition System

Abstract

Automation has become much sought-after knowledge in the modern world. The primary objective of trying to get it enhanced is to save time and energy and make the world a comfortable place for us to live. Automated recognition of currency notes is a branch of knowledge which is going through an evolutionary process and it is being developed in many countries by way of research and experiments. As a result, in many parts of the world, it has gained ground as an integral part of a service extended to activities such as banking sector, shopping malls, vending machines, railway ticket counters, filling stations and currency exchange services etc. When compared with many other countries in this region, Sri Lanka has made significant strides in certain service oriented activities. However, when it comes to automated means of recognizing its own currency notes, Sri Lanka lags far behind many other countries in the world. This anomaly can be attributed to insignificant research and development work done or being done on this particular branch of knowledge by both public and private sectors of the country. This research is an attempt to recognize a new currency note series issued in February 2011 by the Central Bank of Sri Lanka. Accordingly, this system is designed to recognize any currency note of the relevant series including those torn or worn out from either direction. RGB value analysis, image processing technique and optical currency recognition have been used to identify unique features of the relevant currency notes such as their distance of diagonals, color, serial prefix, blind recognition values and artistic impressions.