This application is related to an internal project, called “Coin Recognition Project” behalf of Telecommunication Systems Institute. The intention of the project was to examine the possibility of applying useful image pattern recognition applications in embedded devices with the capabilities of the modern mobile phone devices, known as smart phones. The application was developed in that scope tries to recognize the identity of the coins that a user can capture by its smart phone camera. The main constraints need to be overcome for this purpose is low frequency processors and the lack of memory available in this kind of devices. The development of the application was the main concept of this research. This is the basis for experimental processes and the improvement proposals.
The Coin recognition Project iOS Application offers a user the ability of recognizing coins identity by photographing them or grabbing its images from the device’s photo album. The recognition comes by matching the image the user provides with a total number of coins images saved in applications database. The application keeps important samples of different kind of coins and tries to detect them inside the user’s photo using the normalized cross correlation pattern recognition algorithm. The high correlation results give the suspicious of image matching.
As the application is developed for experimental purposes, it is released with an editable database. This means that the user is free to add or remove coins and its samples from the database. The user can add his own image samples and experiment himself. The application is also designed to be used for any image pattern recognition experiments using the “normalized cross correlation” algorithm and is not limited in coin recognition.