Counterfeit currency is a major issue affecting the economy of many countries. Traditional methods of detecting fake currency rely mainly on manual inspection or specialized hardware devices, which may not always be accurate or accessible. This project proposes a machine learning-based system to detect fake currency notes using image processing techniques. The system analyzes images of currency notes and extracts important features such as texture, color patterns, serial numbers, and security marks. Machine learning algorithms are then used to classify the currency as genuine or fake. By automating the detection process, the proposed system helps reduce human errors and improves the accuracy and efficiency of counterfeit currency identification.