Research on Print Quality Assessment and Identification: Evaluation of Print Edge Roughness

Wide spread of printers and computers have led to increasing use of print documents in people´s common life. While mass print, automatic print quality evaluation is necessary in the fast print process to avoid the quality defect immediately when it appears. Also, in many cases print materials are direct accessories to many criminal and terrorist acts, for example forge contract, hence print identification has become more importance. In order to evaluate the print quality automatically or identify the source of the printouts intelligently, this study researches on the feature extraction on some print edges and evaluates the edge roughness by local high order correlation (LHOC) method. A magnified image acquisition system is developed to capture the edge roughness clearly and then correlation spectrums of each order are employed to describe the roughness degree. The experimental results are then presented and analyzed, showing that the results are promising and convinced.