Abstract
A method is described to extract an index from a binary character image for database storage and retrieval. The index is hierarchical in nature, and different parts can retrieve character image clusters of different sizes. A hierarchy of Kohonen neural networks is used to generate the clusters on which the indexing method is based. The application is the identification of the font from a sample of printed characters. This is particularly important for forensic document examiners who may have to match a document of unknown origin with the equipment that may have produced it. A computerised system is expected to considerably reduce the time required for each document search.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.References
Gupta SK, Nehra TR, Bami HL. An assessment of the Interpol typewriter classification system.J Police Science and Administration 1973;l:409–413
Totty RN, Hall MG, Hardcastle RA, Brown C. A computer based system for the identification of unknown typestyles.Forensic Science Soc 1981;22:65–73
Haas J, Haas B. Atlas of Typewriter Typestyles, Zurich Canton Police, 1987
Badal DZ. Image indexing and neural networks.Neural Network World 1994;3:3–17
Kohonen T. The neural phonetic typewriter.Computer 1988;21:11–22
Augusteijn MF, McCarthy KK. Image indexing applied to character font recognition by means of a Kohonen neural network hierarchy.Intelligent Engineering Systems Through Artificial Neural Networks 1995;5:431–436
Caudell TP, Smith SDG, Escobedo R, Anderson M. NIRS: Large scale ART-1 neural architectures for engineering design retrieval.Neural Networks 1994; 7:1339–1350
Carpenter GA, Grossberg S. A massively parallel architecture for a self-organizing neural pattern recognition machine.Computer Vision, Graphics and Image Processing 1987;37:54–115
Fritzke B. Growing cell structures — a self-organising network for unsupervised and supervised learning.Neural Networks 1994;7:1441–1460
Augusteijn MF, Winterbottom MC. Invariant object recognition using higher-order neural networks, linesegment spectra and multi-resolution training. Under review.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Augusteijn, M.F., Warrender, C.E. Image indexing applied to typewriter font identification. Neural Comput & Applic 4, 209–217 (1996). https://doi.org/10.1007/BF01413819
Issue Date:
DOI: https://doi.org/10.1007/BF01413819