AutoML‐Driven Deep Learning for Fake Currency Recognition
T. Bhaskar, E. Gangadevi · 2025
False copies of genuine items are referred to as counterfeits. The financial industry is still quite vulnerable to counterfeit money. There are numerous detection techniques, but since there are now free picture manipulation programmers, it has become a significant problem in the banking industry. Finding those key regions for evaluation is the first task because there are many important regions that are represented by currencies. Classifiers can determine if the extracted features are real or not. Without classifiers, we may compare the segmented currency image to the region of the original note. Yet, that is insufficient to determine the authenticity of the specific paper currency image. If we segment the crucial areas, the alignment and edges might not match; therefore, phone money note images might be mistaken for the real thing. To obtain better results, the classifiers will process the extracted features. The method for currency recognition using image processing is suggested in this chapter. Three elements, color, size, and texture, are utilized in the recognition process with this technique. This technology makes it simpler to check the value of paper currency at anytime and anyplace, and it uses the CNN (Convolutional Neural Network). The integration of AutoML not only accelerates the expansion process but also optimizes the performance of the deep learning models making the system flexible to various currencies and counterfeit systems. This chapter experimented with this approach on all Indian religious groups. The characteristics of paper money are crucial in this recognition.