Image Document Plagiarism Detection Using DCT Based Local Filtering and Binary Code Hashing

Plagiarism refers to the intellectual property of using the others work in the field of the research. Since easy access to the wide data has become critical problem for researchers, publishers, educational institution. Data is available from the web, large databases and telecommunication network. This paper discusses the technique of the plagiarism for the detection of the copied data in a document image. The approach presented in the paper uses DCT filtering as a source for match study, because each text and image pattern grouped together will yield different frequency pattern. Based on this frequency similarity the detection system finds copy pasted portions of the documents. Performance of system is analyzed on the basis of number of copy paste detected regions and percentage of plagiarism match in scanned content.