Source color laser printer identification using discrete wavelet transform and feature selection algorithms

It has lately become an important research area of digital forensics to identify the characteristics and the originality of the digital devices. This paper presents a novel method to identify source color laser printer by using discrete wavelet transform and feature selection algorithms. To explore the relationship between color laser printers and color images obtained by scanning printed documents, the proposed approach utilizes image processing techniques and data exploration methods to calculate the features by applying statistical analysis on discrete wavelet transform of scanned images. A model of support vector machines sequentially will be created and trained by using these features to identify the source brand and model of color laser printer. In this study, 10 models of color laser printers and color toner from 6 different brands are used for the experiments. The experimental results reach up to 92.4% identification rate which is significantly superior to the existing known method by 5.5%. The testing performance justifies the proposed identification method is very useful for source color laser printer identification.