Report on Distinguishing Genuine and Forged Banknotes
This study aims to distinguish genuine and counterfeit currencies and enhance the security of banking systems by detecting forged banknotes.
During this project, we only used Python language. Therefore, in case we do not mention, every library and method comes from this language.
The dataset was gathered by UCI in 2012. This dataset is collected from OpenMl and has 1372 samples, no missing value, and two features, as follows, which are continuous real numbers:
• V1: is the variance of Wavelet Transformed image,
• V2: is the skewness of Wavelet transformed image.
V1 - Variance of wavelet transformed image:
This is a continuous real number between -7.0421 and 6.8248 (Table 1). It represents the variability or spread of pixel intensity in the wavelet transform image. A higher variance may indicate greater variation in pixel values in the image.
V2 - skewness of the wavelet transformed image:
This is a real number between -13.7731 and 12.9516 (Table 1), which measures the asymmetry in the intensity distribution of the pixels. Positive skewness indicates a longer or fatter tail on the right side of the distribution.