Authenticity of money using the method KNN (K-Nearest Neighbor) and CNN (Convolutional Neural Network)
Cahya Rahmad, Erfan Rohadi, Rina Lusiana · IOP Conference Series Materials Science and Engineering · 2021
Abstract Nowadays, the circulation of counterfeit money is significantly increased, encouraging to conduct research related to genuine real money detection or counterfeit money based on digital imagery. Technological sophistication one of the quality of printers whose ink is very good and can print money like the original makes the layman should be more wary of money ownership. In this research conducted the authenticity of money using the method KNN (K-Nearest Neighbor) and CNN (Convolutional Neural Network). Accuracy KNN method is 87,75%. While the detection accuracy used by CNN is 96,67%. The results obtained from these 2 methods can still be improved with advanced research namely with pre production on the set and the image used. The data set used has the same exposure level, image capture angle and image size.