A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area
碩士 === 長庚大學 === 電機工程學系 === 102 === Due to the rapid development of consumer electronics, video camera is very suitable for installing on the vehicle as a vision assistance device. However, the real traffic condition is much more complex. All of the factors like weather, driving speed, background, an...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Published: |
2014
|
Online Access: | http://ndltd.ncl.edu.tw/handle/24400982694450600238 |
id |
ndltd-TW-102CGU05442029 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-102CGU054420292015-10-14T00:18:19Z http://ndltd.ncl.edu.tw/handle/24400982694450600238 A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area 車輛側邊盲區之障礙物偵測與警示系統 Kuo Fang Huang 黃國芳 碩士 長庚大學 電機工程學系 102 Due to the rapid development of consumer electronics, video camera is very suitable for installing on the vehicle as a vision assistance device. However, the real traffic condition is much more complex. All of the factors like weather, driving speed, background, and target could affect the accuracy of visual recognition. According to past studies, some researches were focusing on the front or rear car detection in nighttime or studying the pedestrian detection, and others were considering how to use the onboard camera to tell the driver if it’s raining. But only a few studies covered all the different weather conditions and almost none of studies had ever paid attention to the safety at vehicle lateral blind spot area. So, this thesis proposes a pedestrian, bike rider and vehicle detection system for lateral blind spot area. This scheme is robust under different weather conditions and driving speeds. The histogram of gradient (HOG) and support vector machine (SVM) methods are used for pedestrian and bike rider detection, while the image subtraction, edge detection, and wheel detection methods are applied for daytime vehicle detection and the headlights detection for nighttime vehicle detection. Moreover, this study combines the rain and brightness detections to overcome the different weather situations. It applies specific algorithm according to the real weather condition to achieve the same good detection result in nighttime and raining day as daytime. Experimental results show that the proposed system can efficiently detect pedestrian, bike rider and vehicle under various scenarios. J. D. Lee 李建德 2014 學位論文 ; thesis 128 |
collection |
NDLTD |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 長庚大學 === 電機工程學系 === 102 === Due to the rapid development of consumer electronics, video camera is very suitable for installing on the vehicle as a vision assistance device. However, the real traffic condition is much more complex. All of the factors like weather, driving speed, background, and target could affect the accuracy of visual recognition. According to past studies, some researches were focusing on the front or rear car detection in nighttime or studying the pedestrian detection, and others were considering how to use the onboard camera to tell the driver if it’s raining. But only a few studies covered all the different weather conditions and almost none of studies had ever paid attention to the safety at vehicle lateral blind spot area. So, this thesis proposes a pedestrian, bike rider and vehicle detection system for lateral blind spot area. This scheme is robust under different weather conditions and driving speeds. The histogram of gradient (HOG) and support vector machine (SVM) methods are used for pedestrian and bike rider detection, while the image subtraction, edge detection, and wheel detection methods are applied for daytime vehicle detection and the headlights detection for nighttime vehicle detection. Moreover, this study combines the rain and brightness detections to overcome the different weather situations. It applies specific algorithm according to the real weather condition to achieve the same good detection result in nighttime and raining day as daytime. Experimental results show that the proposed system can efficiently detect pedestrian, bike rider and vehicle under various scenarios.
|
author2 |
J. D. Lee |
author_facet |
J. D. Lee Kuo Fang Huang 黃國芳 |
author |
Kuo Fang Huang 黃國芳 |
spellingShingle |
Kuo Fang Huang 黃國芳 A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
author_sort |
Kuo Fang Huang |
title |
A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
title_short |
A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
title_full |
A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
title_fullStr |
A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
title_full_unstemmed |
A Warning System for Obstacle Detection at Vehicle Lateral Blind Spot Area |
title_sort |
warning system for obstacle detection at vehicle lateral blind spot area |
publishDate |
2014 |
url |
http://ndltd.ncl.edu.tw/handle/24400982694450600238 |
work_keys_str_mv |
AT kuofanghuang awarningsystemforobstacledetectionatvehiclelateralblindspotarea AT huángguófāng awarningsystemforobstacledetectionatvehiclelateralblindspotarea AT kuofanghuang chēliàngcèbiānmángqūzhīzhàngàiwùzhēncèyǔjǐngshìxìtǒng AT huángguófāng chēliàngcèbiānmángqūzhīzhàngàiwùzhēncèyǔjǐngshìxìtǒng AT kuofanghuang warningsystemforobstacledetectionatvehiclelateralblindspotarea AT huángguófāng warningsystemforobstacledetectionatvehiclelateralblindspotarea |
_version_ |
1718088440486035456 |