Fragments of documents are common subjects in forensic analysis of questioned documents. Analysis of torn documents is more challenging owing to sparse data content. A document forensic expert often finds it difficult to extract relevant information and clues when he/she receives plenty of document fragments for forensic analysis from a crime scene. The forensic expert may get confused amidst this pile of document fragments, and might overlook evidences in this huge pool of data. This dissertation aims to help combat this problem by studying scientific methodologies that can narrow down the search space of a forensic expert.
For reliable forensic analysis there’s a need to sort document fragments with similar content types. The thesis explores different computational techniques to sort out similar type document fragments from a heap of documents. Automatic sorting of document fragments demands execution of the following: text/graphics segmentation, segmentation of text type (printed/handwritten), script identification of text, identification of the writer, identifying the font of the printed text. Adopting various image processing and pattern recognition techniques certain methodologies are proposed for accomplishing such tasks.
First external opponent: Prof. Dr. Volker Märgner, Academic Director, Institute for Communications Technology, Braunschweig Technical University, Germany
Second external opponent: Prof. Dr. Giuseppe Pirlo, Dipartimento di Informatica,Università degli Studi di Bari, Italy
Internal opponent: Associate Prof. Simon McCallum, Faculty of Computer Science and Media Technology, Gjøvik University College
Head of the committee: Prof. Dr. Patrick Bours, Faculty of Computer Science and Media Technology, Gjøvik University College
The candidate's principal supervisor: Professor Katrin Franke, IMT, GUC.
The secondary supervisors: Professor Dr. Slobodan Petrovic, IMT, GUC, and Professor Umapada Pal, Indian Statistical Institute, Computer Vision and Pattern Recognition Unit, Kolcata, India.
Dean Nils Kalstad Svendsen conducted the public defense.