Rendusara, Resqi Abdurrazzaaq Putra (2022) Sistem Lane Keeping Pada Mobil Otonom Menggunakan Convolutional Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember Surabaya.
Text
07111840000106-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 September 2024. Download (4MB) | Request a copy |
Abstract
Pada saat mengendarai mobil, pengemudi dituntut untuk dapat fokus ke depan serta menjaga mobil tetap bergerak pada lajurnya. Namun dikarenakan berbagai faktor seperti kurangnya konsentrasi, kelelahan, dan sebab lainnya, perpindahan lajur tanpa sadar kerap terjadi. Hal ini meningkatkan risiko terjadinya kecelakaan yang dapat membahayakan diri sendiri, penumpang, dan orang lain. Maka dari itu, diperlukan fitur pada mobil agar dapat membantu manusia menjaga kendaraan tetap pada lajurnya, fitur ini disebut dengan Lane Keeping Assistance System (LKAS). Pendekatan penelitian LKAS konvensional biasanya dilakukan secara terpisah sehingga proses yang ditempuh menjadi panjang dan membutuhkan banyak pengaturan secara manual. Salah satu metode yang digunakan untuk mengatasi permasalahan ini adalah melalui pendekatan Convolutional Neural Network (CNN) secara ujung-ke-ujung. Penelitian menunjukkan bahwa model LKAS yang dibuat dengan CNN mampu mencegah mobil keluar dari lajurnya pada jalan lurus, dan belok, dan belok dengan putaran U di kecepatan 5, 6, dan 7 meter/detik. Hasil penelitian ini diharapkan menjadi metode alternatif LKAS pada mobil otonom
========================================================================================================================================
When driving a car, the driver ability to focus on the road is needed to be able to keep a car stay in its lane. But sometimes due to lack of concentration, fatigueness, and many other factors, unintended lane changing happened. This increased the risk of accidents that can endanger the driver, passengers, and others. Therefore, a feature needed on the car in order to help humans keep the vehicle in the lane, this feature is called Lane Keeping Assistance System (LKAS). The traditional approach usually carried out separately so that the process taken is long and requires a lot of manual settings. One of the methods used to overcome this problem is through an end-to-end Convolutional Neural Network (CNN) approach. The results of this study are expected to be an alternative method of LKAS on autonomous cars. This research shows that LKAS that made with CNN is able to prevent cars from getting out of their lane on straight road and non-extreme turn at 5, 6, and 7 m/s speed
Item Type: | Thesis (Other) |
---|---|
Additional Information: | RSE 629.893 2 Ren s-1 |
Uncontrolled Keywords: | Mobil Otonom, Lane Keeping System, Deep Learning, Neural Network |
Subjects: | G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis |
Depositing User: | EKO BUDI RAHARJO |
Date Deposited: | 06 Dec 2022 07:16 |
Last Modified: | 06 Dec 2022 07:16 |
URI: | http://repository.its.ac.id/id/eprint/95165 |
Actions (login required)
View Item |