Portable Kontrol Kemudi Otomatis Berbasis Kamera Untuk Mobil Semi Otonom

Dharma, Krisna Pramdya (2025) Portable Kontrol Kemudi Otomatis Berbasis Kamera Untuk Mobil Semi Otonom. Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 6022221048-Master_Thesis.pdf] Text
6022221048-Master_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (19MB) | Request a copy

Abstract

Pendeteksian jalan merupakan komponen utama dalam sistem kendaraan otonom agar kendaraan mampu mengenali dan mengikuti jalur yang tersedia tanpa campur tangan pengemudi. Penelitian ini merupakan bagian dari pengembangan sistem mobil cerdas ICar ITS, dengan mengusulkan pendekatan berbasis DeepLabV3+ yang menggunakan dua backbone, yaitu MobileNetV3-Large untuk segmentasi citra jalan dan Multilayer Perceptron (MLP) untuk regresi sudut kemudi. Sistem ini menerima satu masukan berupa citra RGB malam hari dan secara langsung menghasilkan dua keluaran, yaitu peta segmentasi dan nilai sudut kemudi dalam derajat. Pengambilan citra dilakukan menggunakan kamera yang dipasang pada ketinggian 3 meter dengan sudut kemiringan 20 derajat, dan resolusi citra diperkecil dari 1280×720 piksel menjadi 256×256 piksel. Dataset yang digunakan terdiri atas 2.489 gambar, yang terbagi menjadi 1.742 data pelatihan, 497 data validasi, dan 250 data pengujian. Nilai sudut kemudi yang dihasilkan digunakan sebagai dasar pengaturan pedal gas dan pedal rem pada sistem kemudi elektronik ICar ITS. Hasil evaluasi menunjukkan bahwa akurasi segmentasi mencapai Piksel Akurasi 99,39%, mIoU 65,97%, dan Mean F1-score 66,33%. Untuk regresi sudut kemudi diperoleh MSE 0,6059, RMSE 0,7784, dan MAE 0,4543 derajat. Sistem ini mampu beroperasi dengan kecepatan rata-rata 23,7 FPS dan waktu proses sebesar 35,3 mil detik per citra. Metode ini terbukti efektif untuk prediksi sudut kemudi berbasis visual pada malam hari dan mendukung kemampuan navigasi otonom kendaraan ICar ITS.
======================================================================================================================================
Road detection is a key component in autonomous vehicle systems, enabling the vehicle to recognize and follow available paths without human intervention. This study is part of the development of the intelligent vehicle system ICar ITS, proposing a DeepLabV3+-based approach utilizing two backbones: MobileNetV3-Large for road image segmentation and a Multilayer Perceptron (MLP) for steering angle regression. The system takes a single input in the form of an RGB night-time image and directly produces two outputs: a segmentation map and a predicted steering angle in degrees. Image acquisition is conducted using a camera mounted at a height of 3 meters with a 20-degree tilt angle, and image resolution is reduced from 1280×720 pixels to 256×256 pixels. The dataset consists of 2,489 images, divided into 1,742 training, 497 validation, and 250 test samples. The predicted steering angle values are used as the basis for controlling the throttle and brake pedals within the ICar ITS electronic steering system. Evaluation results show that segmentation accuracy reached Pixel Accuracy of 99.39%, mIoU of 65.97%, and Mean F1-score of 66.33%. For steering angle regression, the model achieved MSE of 0.6059, RMSE of 0.7784, and MAE of 0.4543 degrees. The system operates at an average speed of 23.7 FPS with a processing time of 35.3 milliseconds per image. This method proves effective for visual-based steering angle prediction at night and supports the autonomous navigation capability of the ICar ITS vehicle.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Intelligent Car, Automatic Driving, Segmentasi Citra, Golf Car, Steering Control, Intelligent Car, Automatic Driving, Image Segmentation, Golf Car, Steering Control
Subjects: Q Science > QA Mathematics > QA336 Artificial Intelligence
Q Science > QA Mathematics > QA76.6 Computer programming.
Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Q Science > QA Mathematics > QA76.9 Computer algorithms. Virtual Reality. Computer simulation.
Q Science > QA Mathematics > QA9.58 Algorithms
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
T Technology > TJ Mechanical engineering and machinery > TJ211.415 Mobile robots
T Technology > TJ Mechanical engineering and machinery > TJ217.6 Predictive Control
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7872.F5 Filters (Electric)
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7888.3 Digital computers
T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.5 Motor vehicles Driving
T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.8 Vehicles, Remotely piloted. Autonomous vehicles.
T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL220 Electric vehicles and their batteries, etc.
T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL521.3 Automatic Control
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User: Krisna Pramudya Dharma
Date Deposited: 21 Jul 2025 06:44
Last Modified: 21 Jul 2025 06:44
URI: http://repository.its.ac.id/id/eprint/120284

Actions (login required)

View Item View Item