A new statistical classifier for hand-written character recognition is presented. The system features a preprocessing phase for image normalization and a distance transform applied to the normalized image, which converts a B/W picture into a gray scale one. A k-Nearest-Neighbor classifier follows, based on the distance transform and a suitable metric. The system has an accuracy of 98.96% when applied to the U.S. Post Office ZIP code database, at 0.98% error rate
Microelectronics Group Home Page | Staff |
Research Activities | Teaching Activity |
Publications | Local Stuff |
DEIS Home Page | University of Bologna Home Page |