High-Speed Recognition of Handwritten Amounts on Italian Checks

This chapter describes an Italian bank check processing system. The system is made up of several processing modules, including those for data acquisition, image preprocessing, character center detection, character recognition, courtesy amount recognition, and legal amount recognition. The chapter describes HACRE, a high-speed system for real-time automatic recognition of handwritten Italian checks. The amount on Italian checks consists of only one word, thus no phrase segmentation is required. The amount is written twice: the legal amount, and the courtesy amount. An additional feature of HACRE is its ability to recognize a set of so-called pseudo-characters, which mostly coincide with traditional characters, except for some differences which were explicitly introduced to improve recognition performance. The character recognizer recognizes each individual pseudo-character, using a hybrid approach, which mixes feature-based and neural recognizers. The clustering subsystem aims to group the centers detected by the centering detector into as many clusters as there are pseudo-characters in the courtesy amount.

Paper

The full text of this publication is not hosted on 44B due to licensing.

Read it at OpenAlex

Similar papers

© 2026 NYSGPT2525 LLC