Download
PDF 123.73 KB
master's thesis
26   3
The method for vehicle license plate character recognition
Josip Posavi (2015)
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Software Engineering
Chair of Visual Computing
Metadata
TitleMetoda za prepoznavanje znakova s registarskih oznaka automobila
AuthorJosip Posavi
Mentor(s)Irena Galić (thesis advisor)
Krešimir Romić (thesis advisor)
Abstract
U radu su razvijena tri algoritma. Prvi algoritam korišten je za segmentaciju registarske oznake s slike vozila u boji. Algoritam je podijeljen na tri glavna procesa. U prvom se procesu provodi predprocesiranje ulazne slike. Slijedi proces binarizacije Otsu metodom. Nakon uspješnog odvajanja područja interesa od ostatka slike provodi se morfološka obrada slike. Slika segmentirane registarske oznake šalje se algoritmu za segmentaciju znakova. Segmentirani znakovi spremaju se u unaprijed definirano polje, te se šalju algoritmu za optičko prepoznavanje znakova. Znakovi se prepoznavanju s obzirom na strukturne značajke znaka kao što u broj i položaj petlji u znaku. Algoritmom je postignuta uspješnost izvođenja od 90,38% uz prosječno vrijeme izvršavanja od 1,458 sekunde
KeywordsLicense plate segmentation binarization character segmentation optical character recognition structural features
Parallel title (English)The method for vehicle license plate character recognition
Committee MembersIrena Galić (committee chairperson)
Časlav Livada (committee member)
Krešimir Romić (committee member)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Lower level organizational unitsDepartment of Software Engineering
Chair of Visual Computing
PlaceOsijek
StateCroatia
Scientific field, discipline, subdisciplineTECHNICAL SCIENCES
Electrical Engineering
Telecommunications and Informatics
Study programme typeuniversity
Study levelgraduate
Study programmeGraduate University Study Programme in Electrical Engineering; branch: Communications and Informatics, Power Engineering
Study specializationbranch: Communications and Informatics
Academic title abbreviationmag.ing.el.
Genremaster's thesis
Language Croatian
Defense date2015-10-01
Parallel abstract (English)
This paper describes the development of three algorithms. The first algorithm is used for license plate segmentation from the color image of the vehicle. The algorithm is divided into three main processes. The first process carries out the pre-processing of the input image. The next process carries out the Threshold using the Otsu method. The last process does the morphological processing of the binary image. The segmented license plate is sent as parameter to the character segmentation algorithm. Segmented characters are saved in the pre-defined array and then sent as a parameter to the optical character recognition algorithm. The algorithm recognizes characters based on their structural features like the number and position of loops in the character. The algorithm achieves a success rate of 90,38 % with an average execution time of 1,453 seconds
Parallel keywords (Croatian)Segmentacija registarske oznake binarizacija segmentacija znakova optičko prepoznavanje znakova strukturalne značajke
Resource typetext
Access conditionOpen access
Terms of usehttp://rightsstatements.org/vocab/InC/1.0/
URN:NBNhttps://urn.nsk.hr/urn:nbn:hr:200:526401
CommitterAnka Ovničević