The objective of this research is to achieve a high accuracy number-plate recognition system using image processing and neural network techniques. The system's main parts are image acquisition and image analysis; image acquisition to photo-log a passing vehicle number-plate, and image analysis to read number-plates. This paper focus on the image analysis side of the system. The process begins with the capturing of a video image of a number-plate and adaptively threshold the grey-level image into binary image. Then, a number-plate location task scans the image looking for the area of the number-plate and outputs its coordinates, next segments it to identify the location of the characters included. After that, each character is scaled to fit a 16 x 16 pixel image and a line-thinning algorithm is applied. Finally, there is character reading to recognize or classify the 6 x 16 pixel thinned character image, using neural network techniques.
Computer vision application to automatic number-plate recognition
Einsatz von Computersichtgeräten zur automatischen Nummernschild-Erkennung
1993
8 Seiten, 4 Bilder, 8 Quellen
Aufsatz (Konferenz)
Englisch
Computer vision application to automatic number-plate recognition
Kraftfahrwesen | 1993
|Computer Vision Application to Automatic Number Plate Recognition
British Library Conference Proceedings | 1995
|93AT047 Computer Vision Application to Automatic Number Plate Recognition
British Library Conference Proceedings | 1993
|Automatic number plate recognition
Tema Archiv | 1998
|Automatic number-plate recognition
Tema Archiv | 1990
|