Advancement of Technology has replaced humans in almost every field with machines. By introducing machines, banking automation has reduced human workload. More care is required to handle currency, which is reduced by automation of banking. The identification of the currency value is hard when currency notes are blurry or damaged. Complex designs are included to enhance security of currency. This makes the task of currency recognition very difficult. To correctly recognize a currency it is very significant to choose the good features and suitable algorithm. In proposed method, Canny Edge Detector is used for segmentation and for classification, NN pattern recognition tool is used which gives 95.6% accuracy.