以電腦視覺為基礎之夜間車輛防撞系統

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：13

、訪客IP：18.118.200.197

姓名

盧映宇(Yin-Yu Lu) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

以電腦視覺為基礎之夜間車輛防撞系統
(A Vision-based Preventing System for Car Collision at Night)

相關論文

★ 使用視位與語音生物特徵作即時線上身分辨識	★ 以影像為基礎之SMD包裝料帶對位系統
★ 手持式行動裝置內容偽變造偵測暨刪除內容資料復原的研究	★ 基於SIFT演算法進行車牌認證
★ 基於動態線性決策函數之區域圖樣特徵於人臉辨識應用	★ 基於GPU的SAR資料庫模擬器：SAR回波訊號與影像資料庫平行化架構 (PASSED)
★ 利用掌紋作個人身份之確認	★ 利用色彩統計與鏡頭運鏡方式作視訊索引
★ 利用欄位群聚特徵和四個方向相鄰樹作表格文件分類	★ 筆劃特徵用於離線中文字的辨認
★ 利用可調式區塊比對並結合多圖像資訊之影像運動向量估測	★ 彩色影像分析及其應用於色彩量化影像搜尋及人臉偵測
★ 中英文名片商標的擷取及辨識	★ 利用虛筆資訊特徵作中文簽名確認
★ 基於三角幾何學及顏色特徵作人臉偵測、人臉角度分類與人臉辨識	★ 一個以膚色為基礎之互補人臉偵測策略

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

保持安全距離在道路駕駛上是十分重要的，保持和前車的安全距離可以有效降低事故發生的機率，這篇論文中提出了一個有效率的距離量測方法，並且只使用了單一攝影機，有別於一般雙眼立體視覺的架構。
本篇論文共分成四個模組，分別為車燈偵測模組、距離量測模組、車牌偵測模組以及決策模組，透過四個模組的分工，我們便可以知道前車的距離，進一步可以對駕駛人做出警告，避免對前車的碰撞。
在車燈偵測模組中，文中以車燈的亮度及顏色特徵作為車燈偵測的基礎，可以在很短的時間內找到車子的兩個車尾燈，利用車尾燈的寬度帶入距離量測模組中，便可以得到與前車的距離，在距離沒有碰撞的危險時我們使用攝影機變焦的功能，取得車牌的影像，利用車牌偵測模組找到車牌的切確位置，並透過車牌與車燈的比例關係對於距離量測所使用的參數做出校正，已達到更精確的防碰撞效果，在決策模組中，以簡單的物理公式給予使用者適當的輔助，以達成安全駕駛的目標。

摘要(英)

Keeping a safe distance away from the frontal car is an important issue for car accident prevention. This thesis presents a practical distance measuring method in nighttime using a single CCD camera. The proposed system consists of three modules including taillight detection, license-plate detection, and distance measurement of which estimating the distance from the frontal car during driving. Firstly, the two taillights of a car are detected and extracted to be the salient features. Based on the proportionality of similar triangles, the distance between the CCD camera and the frontal car can be estimated. In addition, the license plate is detected and the measuring parameters are refined for accuracy enhancement. As a result, less processing time and high accuracy rate can be achieved by using the proposed method.
In taillight detection module, the color and intensity are used to serve as the features. In a special color space, the taillight can be found easily. The second part in the system is the distance measuring module. In this module, we propose a special measuring system in which the real distance can be calculated by just using a single image. In the third module, an Adaboost license plate detector is employed. The license plate can be successfully detected by using Haar function and the system parameter can be automatically improved to uplift the accuracy.
Experimental results demonstrate that our proposed system can indeed achieve the goal of car collision prevention during nighttime.

關鍵字(中)

★ 距離量測

關鍵字(英)

★ distance measuring

論文目次

Abstract.......................................................................................................i
摘要...........................................................................................................iii
誌謝...........................................................................................................iv
目錄.............................................................................................................v
圖目錄......................................................................................................vii
表目錄.......................................................................................................ix
第一章緒論...............................................................................................1
1.1 動機...............................................................................................1
1.2 相關研究.......................................................................................2
1.3 系統流程.......................................................................................2
1.4 論文架構.......................................................................................3
第二章影像式距離量測..........................................................................5
2.1影像式量測系統回顧....................................................................5
2.1.1三角量測法..........................................................................6
2.1.2 雷射光束平行量測法.........................................................8
2.2量測參數視角θ的取得.............................................................10
2.3 改良式平行雷射法....................................................................12
2.4系統攝影機的架設.....................................................................14
第三章類神經網路與車燈偵測...........................................................16
3.1 車燈的特性與特徵的抽取........................................................16
3.1.1 YCbCr與RGB色彩空間................................................16
3.1.2 在不同色彩空間的車燈資訊..........................................19
3.1.2.1 車燈周圍資訊特徵.......................................................20
3.2 倒傳遞類神經網路與連通元件演算法...................................21
3.2.1倒傳遞類神經網路[20]....................................................21
3.2.2連通元件演算法(connected component).........................24
3.3車燈偵測.....................................................................................27
3.4 車燈位置、相機架設與距離量測...........................................31
3.4.1 車燈位置所產生的差異..................................................31
v
3.4.2 相機傾斜..........................................................................32
第四章車牌偵測與參數校正...............................................................35
4.1使用Adaboost做車牌偵測的原因...........................................35
4.2 Adaboost物體偵測....................................................................36
4.2.1 使用Haar Wavelet做特徵抽取......................................36
4.2.2 Integral image...................................................................37
4.2.3 Adaboost訓練流程...........................................................39
4.2.4偵測流程介紹...................................................................41
4.3 系統參數的校正........................................................................42
第五章系統架構與決策模組...............................................................44
5.1系統流程說明.............................................................................44
5.2決策模組.....................................................................................45
第六章實驗與討論...............................................................................49
6.1 實驗環境....................................................................................49
6.2 車燈偵測....................................................................................50
6.3 車牌偵測....................................................................................52
6.4 各模組運算時間........................................................................55
第七章結論與未來工作.......................................................................56
7.1結論.............................................................................................56
7.2 未來工作....................................................................................56
第八章參考文獻....................................................................................58

參考文獻

[1]
K. T. Song, C. H. Chen and C. H. Chiu-Huang, “Design and experimental study of on ultrasonic sensor system for lateral collision avoidance at low speed”, Proc. of 2004 IEEE int. Symposium on Intelligent Vehicles, pp.647-652, 2004.
[2]
Z. Yi, H. Y Khing, C. C. Seng, Z. X. Wei, “Multi-ultrasonic sensor fusion for mobile robots,” Proc. of 2000 IEEE Int. Symposium on Intelligent Vehicles, pp.387-391, 2000.
[3]
N. Tamiya, H. Mandai, T. Fukae, “Optical spread spectrum radar for lateral detection in vehicles,” Proc. of 1996 IEEE Int. Symposium on Spread Spectrum Techniques and Applications, pp.195-198, 1996.
[4]
S. Tanaka, T. Nakagawa, M. Akasu, Y. Fukushima, and W. J. Bracken, “Development of a compact scan laser radar,” in Proc. SAE World Congr., 1999, pp. 101-106.
[5]
K. Hanawa and Y. Sogawa, “Development of stereo image recognition system for active driving assist,” in Proc. IEEE Intelligent vehicles Symposium, 2001. pp. 177-182.
[6]
C. Knoppel, A. Schanz, and B. Michaelis, “Robust vehicle detection at large distance using low resolution cameras,” Proceedings of IEEE Intelligent Vehicles Symposium, pp.267-272, Oct. 2000.
[7]
H.Y. Cheng, B. S. Jeng, P. T. Tseng, and K. C Fan, “Lane Detection With Moving Vehicles in the Traffic Scenes,” IEEE Transaction on Intelligent Transportation System, Vol. 7, NO. 4, Dec. 2006
[8]
Y. U. Yim and S. Y. Oh, “Three-feature based automatic lane detection algorithm (TFALDA) for autonomous driving,” IEEE Transaction on Intelligent Transportation System, Vol. 4, Issue 4, Dec. 2003
[9]
P. Jeong and S. Nedevschi, “Efficient and Robust Classification Method Using Combined Feature Vector for Lane Detection, ” IEEE Transaction on Intelligent Transportation System,” Vol. 15 No. 4, APRIL 200
[10]
E. Dagan, O. Mano, G. P. Stein and A. Shashua, ” 2004 IEEE Intelligent Vehicles Symposium,” University of Parma, Italy June 1447,2004
[11]
G. P. Stein, O. Mano, and A. Shashua, ” 2003 IEEE Intelligent Vehicles Symposium,” Page(s):120 – 125, 9-11 June 2003
[12]
M. C. Lu, W. Y. Wang, and H. H. Lian, “Image-based height measuring system for liquid or particles in tanks,” IEEE International Conference on Networking, Sensing and Control, Vol.1, pp.24-29, 2004.
[13]
M. C. Lu, W. Y. Wang, and C. Y. Chu, “Optical-based distance measuring system (ODMS),” The Eighth International Conference on Automation Technology, pp.282-285, 2005.
[14]
W. Y. Wang, M. C. Lu, T. H. Wang, C. P. Tsai, and Y. Y. Lu, “Image-and robot-based distance and area measuring”,” Proceedings of the 2007 WSEAS Int. Conference on Circuits, Systems, Signal and Telecommunications (CISST’07), Gold Coast, Queensland, Australia, January 17-19, 2007.2007 WSEAS international Conference on (CISST), Gold-Coast, Australia, January 17-19, 2007.
[15]
W. Y. Wang, M. C. Lu, and C. Y. Chu, “Nighttime vehicle distance measuring system(NVDMS),” IEEE Transactions on Circuits and System-II, Volume 54, Jan. 2007.
[16]
T. H. Wang, M. C. Lu, C. C. Hsu, and W. Y. Wang, “A method of distance measurement by digital camera,” Proceedings of 2006 ACAS Automatic Control Conference , St. John’s University, Tamsui, Taiwan, Nov. 10-11, 20032006.
[17]
M. C. Lu, W. Y. Wang, and H. H. Lian, “Image-based distance and area measuring systems,” IEEE Sensors Jornal, vol. 6. no. 2, pp. 495-503, apr.
[18]
H. Zhang, W. Jia, X. He, and Q. Wu, “A fast algorithm for license plate detection in various conditions,” IEEE International conference on Systems, Man, and Cybernetics, Taipei, Taiwan, October 8-11, 2006
[19]
T.-H. Wang, M.-C. Lu, C.-C. Hsu, Y.-Y Lu, and C.-P. Tsai, “Three dimensional distance measurement based on single digital camera,” Proceedings of the 2007 WSEAS Int. Conference on Circuits, Systems, Signal and Telecommunications (CISST’07), Gold Coast, Queensland, Australia, January 17-19, 2007.
[20]
蘇木春, 張孝德, “機器學習：類神經網路、模糊系統以及基因演算法則”, 全華科技圖書股份有限公司, 2003.
[21]
P. Viola and M. Jones, "Rapid object detection using a boosted cascade of simple features," in Proceeding of the IEEE International Conference on Computer Vision and Pattern Recognition, Vol. 1, pp. 1-511 – 1-518, 2001.

指導教授

范國清(Kuo-Chin Fan)

審核日期

2008-7-22

推文