Counterfeit Currency Detection using Deep Convolutional Neural Network
Kiran P. Kamble, Anuthi Bhansali, Pranali Satalgaonkar, Shruti Alagundgi · 2019
Counterfeit money refers to fake or imitation currency that is produced with an idea to deceive. According to recent reports, demonetization led to all-time high inflow of fake notes into banks, resulting in a spike in suspicious transactions. The existing works to detect a counterfeit note are mostly based on image processing techniques. This paper deals with Deep Learning in which a convolution neural network(CNN) model is built with a motive to identify a counterfeit note on handy devices like smart phones, tablets. The model built was trained and tested on a self-generated dataset. Images are acquired using the smart phone camera and fed to the CNN network. The results obtained are encouraging and can be improvised by further research and improvements in the architecture of Deep CNN model. The testing accuracy obtained is about 85.6%, training and the validation accuracy were 98.57% and 96.55% respectively.