Studies about quantitative examination of laser printed documents based on image physical metric

Abstract

In the field of questioned document examinations, the disadvantages of qualitative examination based on expert experience are becoming more and more obvious. In the source determination of laser-printed documents, the quantitative evaluation based on image quality is more and more important. In this paper, the ImageXpert was used to measure 33 characteristic parameters which are used to characterize the characteristics of laser-printed documents. The laser-printed documents were effectively distinguished using the random forest algorithm and the scatter plots. The data of random forest algorithm and scatter plots were derived from the measurement of the ImageXpert. The results showed that it was feasible to determine the source of the laser-printed documents according to the physical metrics of images and lines. The different brands of printers used in the experiment could be distinguished according to 33 characteristic parameters, and most laser printers of the same brand with different models can also be distinguished. The research results lay a certain foundation for the quantitative examination of laser-printed documents, especially in the determination of page-substitution document and content-addition document.