Computer Laser Print Document Identification and Retrieval
计算机激光打印文档鉴别与检索

In this paper laser printer mechanism is researched, and intelligent print document identification and retrieval are achieved by measuring print shape deviation which reflects the intrinsic tolerances of printer device. A novel polar Hausdorff distance is proposed to obtain effective performance of the print shape matching. Print character set including all samples is applied to compare and improve the identification and retrieval accuracy. Experimental results show the approach is suitable for this application in which the accuracy is 90% for retrieval and the minimum error rate is 17.80% for identification.