Source Identification of Documents Based on LOOP Features
Pushpalata Gonasagi, Mallikarjun Hangarge · Advances in systems analysis, software engineering, and high performance computing book series · 2022
Pinpointing the ownership of documents based on printers is a challenging task. However, many methods have been proposed to identify the printers through printed documents. In this task, the chapter explores a method in simple way to identify the laser printer models. The LOOP (local optimal oriented pattern) method is applied to discriminate 10 laser printer models based on the character images. LOOP is an efficient descriptor to differentiate the images. A 10-fold cross-validation technique is used to classify the printers. The classifiers, namely Linear SVM (support vector machine) and Quadratic SVM, are applied to classify the laser printer models based on the documents at a character level. The experimental result shows that the proposed method is robust and outperforms comparable counterparts in the literature survey. The authors have achieved an average accuracy of Linear and Quadratic SVM classifiers of 99.2% and 99.8%, respectively.