., Andrew (2017) Deteksi kecepatan kendaraan berjalan di jalan menggunakan opencv. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5113100032-Undergraduate_Theses.pdf - Published Version Download (1MB) | Preview |
Abstract
Saat ini, di berbagai kota telah dipasang CCTV pada setiap ruas jalan. Dari CCTV, dapat diketahui kondisi lalu lintas, namun tidak dapat diketahui kecepatan setiap kendaraaan. Oleh karena itu, dibuat perangkat lunak yang dapat mendeteksi kecepatan kendaraan di ruas jalan dari video yang diambil oleh CCTV. Tujuan lainnya adalah untuk mengetahui perbedaan hasil deteksi kecepatan dengan berbagai nilai FPS (Frame Per Second).
Input untuk aplikasi ini adalah video (.avi). Pertama, sistem mengambil Region of Interest (ROI). Selanjutnya, sistem melakukan background subtraction, membuat garis awal dan akhir, memperbarui posisi kendaraan, dan menyimpan hasil kecepatan rata-rata kendaraan ke berkas Excel (.xls).
Skenario uji coba dilakukan berdasarkan nilai FPS pada video (30 FPS, 27 FPS, 25 FPS, dan 20 FPS). Setiap skenario terdapat sub-skenario berdasarkan posisi koordinat garis akhir {(296,0); (282,0); (270,0); dan (248,0)}. Pengujian dilakukan 5 kali setiap skenario, lalu dibandingkan dengan hasil sebenarnya untuk mendapatkan nilai error pada sistem. Error terkecil yang dihasilkan sistem sebesar 2,75% dengan posisi koordinat garis akhir di (282,0) pada skenario 30 FPS.
========================================================================== Nowadays, CCTV camera has been installed on every road segment in many cities. From those cameras, people are able to know the traffic condition, but unable to know the speed of each vehicle. Therefore, the author decided to develop a software that can detect vehicle speed on the road from the video taken by CCTV. Another goal is to identify the difference in speed detection results with various FPS values.
The input for this application is a video with .avi format. First, the system take ROI from the video. For the next step, it performs background subtraction, creates the initial and end line, updates the vehicle’s position and stores its average speed to an Excel file (.xls).
The trial scenarios are based on the FPS value of the video (20, 25, 27, and 30 FPS). Each scenario has sub-scenarios based on the position of the end line {(248, 0); (270, 0); (282, 0); and (296, 0)}. Testing is done for each scenario as much as 5 times, then its result is compared with actual result in order to get the system’s error value. The smallest error generated by the system is 2.75% with the end line coordinate position at (282, 0) in the 30 FPS scenario.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | CCTV; OpenCV; ROI; FPS; Speed Detection; Deteksi Kecepatan |
| Subjects: | T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL725.3 Traffic Control |
| Divisions: | Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | Andrew . . |
| Date Deposited: | 14 Aug 2017 05:05 |
| Last Modified: | 05 Mar 2019 03:53 |
| URI: | http://repository.its.ac.id/id/eprint/42693 |
Actions (login required)
 |
View Item |