Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476326Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Siti Norul Huda Sheikh Abdullah , Dr | |
| dc.contributor.author | Pirahansiah Farshid. (P50109) | |
| dc.date.accessioned | 2023-10-06T09:16:30Z | - |
| dc.date.available | 2023-10-06T09:16:30Z | - |
| dc.date.issued | 2011-06-25 | |
| dc.identifier.other | ukmvital:81889 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476326 | - |
| dc.description | Pattern recognition contributes to a wide spectrum of applications, such as, optical character recognition (OCR), biometrics, diagnostic systems and military applications. Nowadays, license plate recognition (LPR) has been one of the most important issues in OCR for identification of motor vehicles for the purpose of law enforcement, border protection, vehicle thefts, automatic toll collection, traffic and speed control, and entrance admission to car parks and visitors premises, security checks, and parking control. This research has studied adaptive threshold segmentation algorithm for image pattern recognition. In this research, Malaysian LPR systems have been studied. Upon separating foreground from background, thresholding has become a critical stage in image processing applications. This research proposed an adaptive single threshold segmentation algorithm and the proposed algorithm has been applied for license plates. This algorithm has been developed by combining two methods, which are multi level and multi threshold techniques that can be applied in the wide range of pattern recognition applications. This algorithm uses Peak Signal to Noise Ratio (PSNR) for finding threshold value from all threshold values. In addition, four comparisons have been conducted in order to evaluate the performance of proposed algorithm compared to other algorithms such as Kittler and Illingworth's MET, Potential difference, and Otsu's method. The proposed thresholding method was tested within the LPR system. In addition, this research presented in our test data that consisting 1216 images of vehicle. From the experiment, Kittler and Illingworth's MET, Potential Difference, Otsu’s, multi threshold, multi level and the proposed methods have been achieved as 90.19%, 95.51%, 58.83%, 99.67% , 94.85% and 90.75% respectively, in license plate detection phase. This algorithm can be further improved to increase the accuracy of segmentation. Furthermore, we have performed a comparative study on the proposed algorithm with printed , hand written images and standard images in order to assess its potential in a broad range of image processing application.,Master | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Dissertations, Academic -- Malaysia | |
| dc.subject | Image Processing | |
| dc.subject | Optical Pattern Recognition | |
| dc.subject | Pattern Recognition Systems | |
| dc.title | Adaptive single thresholding method for image segmentation based on peak signal-to-noise ratio | |
| dc.type | theses | |
| dc.format.pages | 470 | |
| dc.identifier.callno | TA1650.P535 2011 | |
| dc.identifier.barcode | 001541 | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_81889+SOURCE1+SOURCE1.0.PDF Restricted Access | 11.15 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.