<![CDATA[Banknotes that distributed by scammers would causes restlessness in society, including blind people. With impairment vision, blind people would hard to distinguish between genuine and fake money. From that problem there would be a research about nominal and authenticity detection system for blind people. The detection system consists of camera as a sensor device to detect picture from the banknote, followed by ultraviolet lamp to tell about the genuine banknote, and speaker as the output from this system. Output would generate voice as the nominations and tell if it is genuine or fake banknote. Code program on this system were written in Pyhton language with Raspberry Pi hardware, Webcam sensor camera, and ultraviolet lamp. Detecting banknotes would use Histogram of Oriented Gradients method and using K-Nearest Neighbour method to classify banknote. Around 3370 data training were used to detect about authenticity of the banknotes and the detections were tried for 56 times. implementation of K-Nearest Neighbor method using k=3 obtained an accuracy result of 98.21% with an average compute time of 3608 ms.]]>
Copyrights © 2021
