As of today, at least 2.2billion people suffer from one or another visual impairment. India has around 40 million people, who are either blind or visually impaired. They face a lot of challenges regarding the monetary transaction which they require for their daily survival. This research study has proposed an Arduino-based currency detection system to assist them in resolving their issues. The currency note is captured through a camera which is then given to a Machine Learning model. The model is based on convolution neural network which helps in image processing and classification. Various filters like convolution 2D, ReLU, max pooling, flatten, and dense are used for feature extraction of the input image and finally, classification of the input image is done. Lastly, the results are delivered through a speaker.