This paper presents advances in zone classification for printed document image analysis. It firstly introduces entropic heuristic for text separation problem. Then a brief recall on existing texture and geometric discriminant parameters proposed in a previous research is done. Several of them are chosen and modified to perform statistical pattern recognition. For each of these two aspects, experiments are done. A document image database with groundtruth is used. Available results are discussed.
CITATION STYLE
Duong, J., Côté, M., & Emptoz, H. (2002). Feature approach for printed document image analysis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2396, pp. 159–167). Springer Verlag. https://doi.org/10.1007/3-540-70659-3_16
Mendeley helps you to discover research relevant for your work.