Automātiska automašīnu numura zīmju atpazīšana
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
N/A
Abstract
Darbā apskatīti attēlu analīzes paņēmieni, par pamatu ņemot to pielietošanu automašīnu
numura zīmju atpazīšanā. Soļi numura zīmju atpazīšanā sastāv no automašīnu atrašanas attēlā,
numura zīmes atrašanās vietas noteikšanas, un simbolu atpazīšanas uz numura zīmes. Apskatītas arī
vairākas metodes, kas pielietojamas šo soļu veikšanā, lielāko daļu uzmanības veltot dinamisku attēlu
(videoattēlu) analīzē. Darbā apskatīti attēlu analīzes pamati, vairākas vienkāršas metodes attēlu
analīzē (attēlu starpība, šķēlums), kā arī autors piedāvā vairākas metodes (fona atrašana secīgu attēlu
virknē, automašīnu atrašana, ņemot vērā perspektīvu). Apskatītie algoritmi arī izstrādāti, izmantojot
programmēšanas valodu Python un izmantojot attēlu apstrādes bibliotēku PIL (Python Imaging
Library.
There are several methods of image analysis in this work, mainly focusing on car number plate recognition. There are several steps in recognition of car number plates – finding cars in picture, locating number plate, and finally recognition of characters on number plate. Several methods are explained in this work, most attention paying to analysis of dynamic images (video images). Fundamentals of image analysis are explained in more detail, continuing with some basic methods of image analysis (image intersection and subtraction). Author also explains more advanced subjects like background extraction from a sequence of images and locating cars in pictures. Algorithms were implemented in Python programming language using PIL (Python Imaging Library).
There are several methods of image analysis in this work, mainly focusing on car number plate recognition. There are several steps in recognition of car number plates – finding cars in picture, locating number plate, and finally recognition of characters on number plate. Several methods are explained in this work, most attention paying to analysis of dynamic images (video images). Fundamentals of image analysis are explained in more detail, continuing with some basic methods of image analysis (image intersection and subtraction). Author also explains more advanced subjects like background extraction from a sequence of images and locating cars in pictures. Algorithms were implemented in Python programming language using PIL (Python Imaging Library).