IR@KDU Repository

Histogram Based Number Plate Recognition with Template Matching for Automatic Granting

Show simple item record

dc.contributor.author Weerasinghe, LDSB
dc.contributor.author Tennegedara, TGDS
dc.contributor.author Jinasena, TMKK
dc.date.accessioned 2018-05-18T15:09:30Z
dc.date.available 2018-05-18T15:09:30Z
dc.date.issued 2015
dc.identifier.uri http://ir.kdu.ac.lk/handle/345/1047
dc.description Article Full Text en_US
dc.description.abstract Automatic Vehicle number plate recognition (AVNR) is a very popular and highly demanding in transportation industry, law and enforcement such as automating fines for traffic rule violations, automating parking lots, entry and exit in highways, quickly identifying granted vehicles. Most of the AVNRs use image processing techniques such as image enhancement, restoration, segmentations, block-based character recognition, optical character recognition and template matching. This paper presents the design and implementation of such a system to General Sir John Kotelawala Defence University (KDU) to capture and recode vehicle logs. People face difficulties at the gate premises due to the existing manual system. To collect the data, qualitative research techniques such as interviews and observations were used. Once an image captured, median filters were used to remove noises. Sharpen filters used to detect edges. Camera calibration was used to correct the perspective view. Lines were detected using Hough transformation and, rectangles were identified. Histogram processing is used to identify the number plates. Template matching with different letters, numbers, and special character were used to recognize its content. Once a license plate is found, its figures are recognized, displayed on the user interface and checked in the database for grants. Moreover arrival or departure times are being recorded. MATLAB was used to develop the prototype together with MySQL. Two cameras with 720x480 resolution and an i5 laptop with 4GB RAM is used to test the system. Accuracy of the system was over 84%. Accuracy was differing under different muddy conditions. As a quantitative research technique data, such as light conditions, distance from vehicle to camera have collected and analysed. Findings show that some of the factors like non-uniform vehicle number plates, language specific characters such as “Sri” and dash, are challenging in AVNR. Post research questioners have proven that the system is very useful and time saving methods for KDU and also its reliability en_US
dc.language.iso en_US en_US
dc.subject Image Processing en_US
dc.subject Camera Calibration en_US
dc.subject Histogram en_US
dc.title Histogram Based Number Plate Recognition with Template Matching for Automatic Granting en_US
dc.type Article Full Text en_US
dcterms.bibliographicCitation Weerasinghe, L., Tennegedara, T. and Jinasena, T. (2015) ‘Histogram Based Number Plate Recognition with Template Matching for Automatic Granting’, in KDU International Research Symposium Proceedings. General Sir John Kotelawala Defence University, pp. 111–116. Available at: http://ir.kdu.ac.lk/handle/345/1047.
dc.identifier.journal KDU IRC en_US
dc.identifier.pgnos 112-116 en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search IR@KDU


Browse

My Account