Title :
Number plate recognition by hierarchical neural networks
Author_Institution :
Automa-Sistemi di Autom. Ind. s.c.r.l., Genova, Italy
Abstract :
A system for the automatic identification of wagons through number plate optical readings is here described. Techniques based on neural networks are applied to number recognition. The system is composed of several neural networks connected in a cascaded structure to diminish the calculation time and the error rate. Moreover a neural network based procedure for the selection of the more relevant features for the recognition is presented.
Keywords :
image processing; multilayer perceptrons; optical character recognition; railways; automatic identification; calculation time; cascaded structure; error rate; hierarchical neural networks; number plate optical readings; number plate recognition; wagons; Biomedical optical imaging; Character recognition; Error analysis; Image recognition; Neural networks; Optical character recognition software; Optical computing; Rail transportation; Vehicles;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.714139
