Lukas, Kristopher (2018) Rancang Bangun Pengenalan Citra Rambu Lalu Lintas dengan Metode Local Binary Pattern. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07111340000105-Undergraduate_Theses.pdf - Accepted Version Download (4MB) | Preview |
Abstract
Pada Tahun 2013, data kecelakaan yang didapat dari Departemen Perhubungan Indonesia, terdapat setidaknya 8 kecelakaan yang 6 diantaranya disebabkan oleh kesalahan dari manusia. Hal inilah yang menyebabkan pengembangan teknologi pada keselamatan berkendara. Autonomus Vehicle merupakan teknologi yang mengambil alih peran manusia dalam mengemudi kendaraan.
Metode Local Binary Pattern mengubah sebuah citra untuk didapatkan nilai histogram vector yang tidak dipengaruhi oleh intensitas cahaya pada citra. Pada program akan dilakukan pengambilan citra untuk data learning dan pendeteksian rambu dengan Hough Transformation. Pada pengenalan dilakukan komparasi nilai histogram citra dengan data learning menggunakan metode Chi-Square. Dari hasil data yang didapatakan, bahwa pengujian pengenalan sebuah citra dengan menggunakan metode LBP menghasilkan tingkat akurasi pengenalan 96%, dengan menggunakan proses segmentasi sebanyak 100 subdivide.
Pada penelitian tugas akhir ini didapatkan bahwa pengenalan rambu lalu lintas menggunakan metode LBP sangat baik. Dari penelitian ini akan dilakukan pengembangan terhadap pengenalan rambu lalu lintas pada autonomous vehicle, sehingga dapat berkendara di jalan raya tanpa melanggar hukum. ================= There were 8 traffic accident, 6 of them were road coallison caused by human error in 2013, from Indonesia Departemen of Traffic. This kind of accident trigger some concern in safety feature in traffic. Autonomus
Vehcile Technology is one of the technology that can reduce or eliminate people role to drive. In this final project a system has been designed and built to recognize traffic sign using Local Binary Pattern method. This program
consists of learning phase with collecting image for data learning and detect object using Hough Transformation. Second phase is recognition, this consist of LBP procese and compare histogram from input images with data learning using Chi-Sqr. To get better recognition, this method combines with segmentation to get vector histogram. After testing this program, it shows a result that the recognize has 96% accuracy with using 100 segmentations for vector histogram. In this research, recognize the traffic sign using LBP method show
great result. This method doesn’t affect the illumination of image so the image can still be recognized when its night. Further work for this program is new development of LBP method to be use as the recognizer in fast image recognizer and development for traffic sign recognizer for
autonomous vehicle so it can run on traffic.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSE 006.4 Luk r |
| Uncontrolled Keywords: | Local Binary Pattern, Pengenal Objek, OpenCV, Hough Transformation, Local Binary Pattern, Object Recognition, OpenCV, Hough Transformation. |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7870.23 Reliability. Failures T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems |
| Divisions: | Faculty of Electrical Technology > Electrical Engineering > 20201-(S1) Undergraduate Thesis |
| Depositing User: | Kristopher Lukas |
| Date Deposited: | 16 Nov 2018 07:39 |
| Last Modified: | 16 Apr 2021 04:46 |
| URI: | http://repository.its.ac.id/id/eprint/52889 |
Actions (login required)
 |
View Item |