Deteksi Merk dan Tipe Kendaraan Menggunakan Deep Learning Mask R-Cnn

Nafi’i, Mohammad Wahyudi (2019) Deteksi Merk dan Tipe Kendaraan Menggunakan Deep Learning Mask R-Cnn. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 07111750067008-Master_Thesis.pdf Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Deteksi dan klasifikasi kendaraan merupakan bagian yang tidak dapat dipisahkan dari Intellegent Transportation Systems (ITS), berbagai macam penerapan teknologi informasi digunakan agar mampu melakukan deteksi dan klasifikasi kendaraan tersebut, mulai penggunaan ultrasonic sensors, laser scanner, induction loops, magnetic sensor, range sensor, pressure sensor dan kamera CCTV, akan tetapi beredarnya kendaraan dengan desain yang sama dari perusahaan pabrikan yang berbeda membuat klasifikasi kendaraan untuk menentukan merk dan tipe kendaraan menjadi sulit dilakukan.
Pada makalah ini framework deeplearning Mask Regional-Convolutional Neural Network (Mask R-CNN) diusulkan untuk menyelesaikan masalah tersebut. Uji coba telah dilakukan sebanyak tiga kali dengan menggunakan kombinasi dataset dan algoritma deteksi yang berbeda. Agar dapat membedakan mobil dengan kemiripan bentuk yang sama dari perusahaan pabrikan yang berbeda kami menggunakan logo kendaraan sebagai salah satu fitur yang membedakan pabrikan tersebut. Hasil deteksi dan klasifikasi yang terbaik didapatkan pada pelatihan dataset menggunakan iterasi sebanyak 60 epoch, 400 step dengan nilai akurasi mencapai 0.92 dan mAP (Mean Average Precission) sebesar 0.89.
=======================================================================================================================================
Detection and classification of vehicles are inseparable parts of Intelligent Transportation Systems (ITS). Various kinds of information technology applications are used to detect and classify these vehicles, including ultrasonic sensors, laser scanners, induction loops, magnetic sensors, range sensors, pressure sensors, and CCTV cameras. However, the circulation of vehicles with similar designs from different manufacturing companies makes vehicle classification to determine the brands and types of vehicles difficult. In this paper, a deep learning framework, Mask Regional-Convolutional Neural Network (Mask R-CNN), is proposed to solve this problem. Experiments were conducted twice using a combination of different datasets and detection algorithms. To distinguish cars with similar shapes from different manufacturers, vehicle logos were used as one of the features to identify the manufacturer. The best detection and classification results were obtained from dataset training using 60 epochs and 400-step iterations, with an accuracy value of 0.92 and a mean Average Precision (mAP) of 0.89.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Deteksi, Klasifikasi, Logo, Kendaraan, Deep Learning, Mask R-CNN
Subjects:	Q Science > QA Mathematics > QA76.585 Cloud computing. Mobile computing. Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > TA Engineering (General). Civil engineering (General) T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Wahyudi Nafii
Date Deposited:	23 Jul 2026 01:21
Last Modified:	23 Jul 2026 01:21
URI:	http://repository.its.ac.id/id/eprint/65443

Actions (login required)

View Item