Recognition of Vehicle Number Plates

This work deals with problematic from field of artificial intelligence, machine vision and neural networks in construction of an automatic number plate recognition system. (ANPR). This problematic includes mathematical principles and algorithms, which ensure a process of number plate detection, processes of proper characters segmentation, normalization and recognition. Work comparatively deals with methods achieving invariance of systems towards image skew, translations and various light conditions during the capture. Work also contains an implementation of a demonstration model, which is able to proceed these functions over a set of snapshots.