Range estimation in vision-based forward collision warning system

Also available in printed version

Bibliographic Details
Main Author: Mursidi Unir@Hashim
Other Authors: Kamaludin Mohamad Yusof, supervisor
Format: Master's thesis
Language:English
Published: Universiti Teknologi Malaysia 2025
Subjects:
Online Access:https://utmik.utm.my/handle/123456789/115938
Abstract Abstract here
_version_ 1854934138247708672
author Mursidi Unir@Hashim
author2 Kamaludin Mohamad Yusof, supervisor
author_facet Kamaludin Mohamad Yusof, supervisor
Mursidi Unir@Hashim
author_sort Mursidi Unir@Hashim
description Also available in printed version
format Master's thesis
id utm-123456789-115938
institution Universiti Teknologi Malaysia
language English
publishDate 2025
publisher Universiti Teknologi Malaysia
record_format dspace
record_pdf Abstract
spelling utm-123456789-1159382025-08-21T07:36:59Z Range estimation in vision-based forward collision warning system Mursidi Unir@Hashim Kamaludin Mohamad Yusof, supervisor Electrical engineering Also available in printed version Obstacle detection and range estimation are important issues for driver assistance systems. Its have to be performed with high reliability to avoid potential collisions. In most research done to date, the vision-based obstacle detection techniques are highly regarded for this purpose as they require less infrastructure and more economic. However, the performance and robustness of such systems remain challenging to researcher. This research develops and investigates a forward collision warning system using stereo vision sensors (two cameras) which detects vehicles ahead. A depth map is generated from rectified stereo images which contains disparity information which is proportional to the distance of the corresponding world point from the camera. Histograms of oriented gradient (HOG) feature matching is employed to detect the target object. The object distance then can be computed from its 3-D world coordinates. An error compensation model was developed to further improve the accuracy of the distance estimation. The results show that percentage estimation error was reduced after the distance calculation was refined using the proposed error compensation model. The system achieves an average error that is less than 2% at distances up to 15 meters fahmimoksen UTM 87 p. Thesis (Sarjana Kejuruteraan (Elektronik dan Komunikasi)) - Universiti Teknologi Malaysia, 2017 2025-04-22T01:38:58Z 2025-04-22T01:38:58Z 2017 Master's thesis https://utmik.utm.my/handle/123456789/115938 vital:113077 valet-20180718-114541 ENG Closed Access UTM Complete Unpublished Completion application/pdf Universiti Teknologi Malaysia
spellingShingle Electrical engineering
Mursidi Unir@Hashim
Range estimation in vision-based forward collision warning system
thesis_level Master
title Range estimation in vision-based forward collision warning system
title_full Range estimation in vision-based forward collision warning system
title_fullStr Range estimation in vision-based forward collision warning system
title_full_unstemmed Range estimation in vision-based forward collision warning system
title_short Range estimation in vision-based forward collision warning system
title_sort range estimation in vision based forward collision warning system
topic Electrical engineering
url https://utmik.utm.my/handle/123456789/115938
work_keys_str_mv AT mursidiunirhashim rangeestimationinvisionbasedforwardcollisionwarningsystem