relation: https://khub.utp.edu.my/scholars/9652/
title: Vehicle Detection Using Convolutional Neural Network for Autonomous Vehicles
creator: Tarmizi, I.A.
creator: Aziz, A.A.
description: This paper implements a region identification to automatically detect vehicles on the road using Convolutional Neural Network (CNN) for autonomous vehicles. The region identification is essential to ensure the autonomous vehicle can safely navigate without having collision with other vehicles. Existing works highlighted the problem of poor detection in autonomous vehicles (AV) during night scene and in bad weathers. This paper develops a technique to improve the detection of vehicles during night scene (poor lighting condition) and bad weathers using CNN approach. The dataset of 301 vehicles from KITTI Vision, Intercity Roads and Adverse Driving Scenarios (iROADS) and Matlab were used. The CNN was trained with 177 input images in the training part while 124 images were used in the testing part. The simulation results show that the detection accuracy has improved under various weathers. The detection accuracies are 94.3 during a sunny weather, 61.4 during night weather, 73.4 and 98.7 during a snowy weather. In conclusion, the CNN technique has shown a promising result in detecting vehicles during the night scene (poor lighting condition) and bad weathers. © 2018 IEEE.
publisher: Institute of Electrical and Electronics Engineers Inc.
date: 2018
type: Conference or Workshop Item
type: PeerReviewed
identifier:   Tarmizi, I.A. and Aziz, A.A.  (2018) Vehicle Detection Using Convolutional Neural Network for Autonomous Vehicles.  In: UNSPECIFIED.     
relation: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85059763446&doi=10.1109%2fICIAS.2018.8540563&partnerID=40&md5=db4a7ebf541a05b6ce3fd0984fe5e53a
relation: 10.1109/ICIAS.2018.8540563
identifier: 10.1109/ICIAS.2018.8540563