Image processing, verification, and compression techniques are integrated to construct a telephone customer maintenance subsystem in which a legitimate transaction is carried out after the step of verifying the genuineness of the seal imprint on the request form. The key contents includes a distance or similarity metric, matching algorithm, and seal coding. The distance or similarity measure between seals is based on a generalized Euclidean distance which is insensitive to stroke width, minor spatial deformation, broken lines, etc. A block Huffman coding technique is used to reduce the storage of seal imprints. A reasonably good identification rate and 40% storage reduction of sample seals demonstrate that the proposed approach is effective. A 58% storage reduction can be attained when a flexible arithmetic coding method is used
Published in:
Security Technology, 1991. Proceedings. 25th Annual 1991 IEEE International Carnahan Conference on
Date of Conference: 1-3 Oct 1991