1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	Malaysian road tax sticker recognition using optimized color template matching
Authors:	Leila Tahmooresi (P56106)
Supervisor:	Assoc. Prof. Dr. Md. Jan Nordin
Keywords:	Color codes. Image processing-Digital techniques. Transportation Automotive-Taxation.
Issue Date:	5-May-2014
Description:	In Malaysia, vehicle owners are required to pay vehicle road tax, which is paid to the government for a vehicle license (or tax disc), which must be displayed on all vehicle's windscreens on public roads. Currently, checking road tax stickers are done manually by police blocking. Using human resources is not economical, precise and popular. Sticker recognition for human is a repeatable and boring task that it is conducive to some mistakes in checking time. Considering that machine rarely makes a mistake and it is a non-stop and tireless to check road tax stickers in all climatic condition so a robust road tax recognition system is useful and efficient. The main issue of localization is to detect the location of sticker on the windscreen and cropping it. Current localization methods such as SIFT, SURF and template matching, ignore color and subtle object size information when searching the region of interest in an image. Those methods are perfectly performed in gray intensity which has less variety of color changes. For Malaysian road tax localization a new method based on periodic color sweep on original image and analyzing pixels on demand along with three colour localizing approaches are discussed in this study. The proposed system is able to check vehicles road tax sticker in toll stations and critical zones. The proposed system detects and localizes a road tax sticker on the windscreen. This study, proposes a system to localize the location of road tax sticker using color features and template matching. The proposed system uses HSL (Hue, Saturation, Lightness) color space and a description of object model to localize road tax sticker. After localization, all characters on road tax sticker (including road tax expire date and license plate number) are extracted. The proposed system contends 89.16% in correct detection in standard camera position.,Tesis ini tiada Perakuan Tesis Sarjana / Doktor Falsafah"
Pages:	76
Call Number:	TA1637.T337 2014 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.