The ability to recognize currency (both coins and bills) without human input is critical for a number of applications. Probably the most important one is assisting visually impaired people. According to World Health Organization, 284 million people (over 4% of world population) are visually impaired, including 39 million blind. While coins can be potentially recognized based on the shape, bills in the majority of currencies have similar size and often similar colors, which may cause problems even for people with poor vision that are not legally blind. Another aspect of everyday life that could benefit from coin and bill recognition is automated teller machines (ATMs). Presently, most ATMs require a person to place the bills in the envelope for deposition and rely on ATM service personnel for accurate counting. Not only this approach is error-prone but also opens possibility for fraud if the person depositing money intentionally claims larger amount than deposited. These issues can be eliminated by built-in system of banknote recognition.
The aim of this project is to develop a system that could recognize the Pakistani Banknotes using mobile camera on Android mobile using image processing techniques. The ultimate objective is to assist the visually impaired persons to recognize the banknotes. Students would be exploring different features such as SURF features and would come up with set of features that could provide best results both in terms of accuracy and performance.
Discipline:- Computer Science
Hardware/Software/Other Requirements:- Android Platform
Emoticon Emoticon
Note: only a member of this blog may post a comment.