Forensic Analysis of Manuscript Authorship: An Optimized Computational Approach Based on Texture Descriptors
This paper presents an optimized method for establishing the authorship of questioned handwritten documents, on the basis of a forensic analysis and a computational model using texture descriptors. The proposed method uses two classes of texture descriptors: model-based, using fractal geometry, and statistical, using GLCM (Gray-Level Co-occurrence Matrix) and Haralick’s descriptors. The proposed method also uses an SVM (Support Vector Machine) as a classifier and generator of the writer-independent training. The results demonstrate the robustness of the writer-independent obtained from the features by using texture descriptors and robustness in the amount low of samples used as references for comparison and the number of feature used. The results appear promising, in the order of 97.7 %, and are consistent with those obtained in other studies that used the same database.
KeywordsHandwritten Document Classifier Texture Descriptors
- 4.Shomaker, L.: Advances in writer identification and verification. In: Proceedings of the International Conference on Document Analysis and Recognition (ICDAR), Curitiba, Brazil (2007)Google Scholar
- 6.Bush, A., Boles, W., Sridharan, S.: Texture for script identification. IEEE Trans. Pattern Anal. Mach. Intell. 27(11), 1721–1732 (2005)Google Scholar
- 8.Schlapbach, A., Bunke, H.: Using HMM based recognizers for writer identification and verification. In: Ninth International Workshop on Frontiers in Handwriting Recognition, pp. 67–172 (2004)Google Scholar
- 9.Bulacu, M., Shomaker, L.: Writer identification using edge-based directional features. In: Proceedings of 7th International Conference on Document Analysis and Recognition (ICDAR). IEEE Computer Society (2003)Google Scholar
- 10.Cha, S.H.: Use of the distance measures in handwriting analysis. Doctor Theses. State University of New York at Buffalo, EUA, 208 (2001) Google Scholar