| dc.description.abstract | Invented in 1976, Number Plates Recognition (NPR) has since found wide commercial applications, making its research prospects challenging and scientifically interesting. A complete NPR system functions by vz steps, license plate; localization, sizing and orientation, normalization, character recognitions and geometric analysis. This paper is a review of NPR preliminary stages; it explains number plate localization, sizing and orientations as well as normalizations sections of the Number Plates Detection and Recognition-Tanzania Case study. MATLAB R2012b is employed in these processes. The input incorporated includes front and rear photographic images of vehicles, for proximity and simulation purposes the ample angle of image is 90 degree +-15. The captured image is converted to gray scale, binarized and edge detection algorithms are used to enhance edges. The output of this stage provides the input feature extraction, segmentation and recognitions. | en_US |
