臺灣博碩士論文加值系統

本文就嵌入式平台車道偵測系統面臨之問題:ARM(PXA255)在硬體上不支援除法指令和浮點數運算，以嵌入式系統實驗室發展之車道偵測演算法為例，提出一些程式設計優化的方法:盡量減少除法及浮點數的運算。演算法處理的流程為︰載入影像提高對比度二值化車輛頂點搜尋車道特徵點搜尋可能線段連結顯示結果。首先，針對提高對比度副程式中，利用預先建表方式將參考文獻[28]之程式需要執行25600次的除法運算減少到只需執行256次。其次是分別在二值化及車輛頂點搜尋副程式，利用技巧將原本需要做乘法及加法的運算，簡化成只做加法運算即可。在車道特徵點搜尋副程式中，透過資料分析於參考文獻[28]中將需要比對256*100次減少成256*5次，最後在可能線段連結副程式中，評估資料大小的範圍後將原本使用浮點運算改成整數運算，可大幅減少運算時間。在未優化之車道演算法中平台處理時間約為47.5ms/frame，然而經由本文避免除法運算優化及浮點數優化策略，實驗結果顯示優化後車道偵測演算法整體處理時間為28.9ms/frame，優化後處理速度可提升39.28%。

In this thesis, we propose some techniques to overcome the non-supporting division instruction and floating operation problems in an ARM embedded platform (PXA255) for application in the lane detection system. The strategy is to minimize the division instruction and floating operation in the program design. The processing flow of our lane detection algorithms describes as followed: load image  contrast enhancement  binary  find peak points  find edge points  edge connection  line segment combination display result. Firstly, we build a lookup table to reduce the division operation from 25600 to 256 times in the subprogram of contrast enhancement. Secondly, the operation of multiplication and addition was reduced to only addition operation by a predefined address technique in the subprogram of binary and find peak point. Thirdly, in the subprogram of find edge point, we analyze the data group property to reduce the data comparison from 256*100 to 256*5 times. Finally, in the subprogram of edge connection, we use the integer operation to substitute the float-point operation without loss the accuracy. It can greatly reduce the processing time. The lane detection process time is about 47.5 ms/frame before program design optimization. After optimization, the lane detection process time is about 28.9 ms/frame. The enhancement of processing speed is as high as 39%.

目錄
指導教授推薦書 i
口試委員審書 ii
長庚大學博碩士論文著作授權書 iii
致謝 iv
中文摘要 v
ABSTRACT vi
目錄 vii
表目錄 ix
第一章 緒論 1
1.1 前言 1
1.2 文獻回顧 2
第二章 影像前處理 7
2.1 影像處理系統的硬體組成 7
2.2 數位影像的表示法 9
2.3 何謂清晰的影像 9
2.4 增加影像對比度 11
2.5 二值化 14
2.6 動態調整二值化區域 16
第三章 車道線偵測演算法 19
3.1 車輛偵測系統架構[24] 19
3.2 車道特徵點擷取[24] 21
3.3 車道線段偵測[24] 24
3.4 車道標線特徵點聚集[24] 25
3.5 車道線段接合[24] 29
3.6 車道標線邊界篩選[24] 32
3.7 車道接合與選取[24] 34
3.8 減少資料量[23] 35
第四章 實驗平台、分析與結果 36
4.1 平台簡介[28] 36
4.2 車道辨識系統之背景[28] 39
4.3 平台問題[28] 42
4.4 實驗設計[28] 45
4.5 挑戰新平台 49
4.6 優化車道偵測辨識演算法[28] 51
4.6.1 優化強化對比[28] 53
4.6.2 優化二值化[28] 58
4.6.3 優化頂點搜尋[28] 60
4.6.4 優化特徵點聚集[28] 62
4.6.5 優化線段連結[28] 68
4.6.6 優化前後測試結果[28] 70
第五章 結論 72
5.1 結論 72
5.2 綜合比較 74
參考文獻 76

表目錄
表1-1 行駛速度下車道辨識所需距離 2
表4-1 PXA255硬體規格 37
表4-2 PXA255使用的個人電腦規格與編譯環境 39
表4-3 History Summary 42
表4-4愛星AC-511 Camera規格 45
表4-5 SAMSUNG6410L硬體規格 47
表4-6 新灰階Table表 55
表4-7 Find Edge Point比較表 67
表4-8整數與浮點數值域比較表 70
表4-9 Optimization Summary 71
表5-1效能比較表 75

圖目錄
圖2-1影像處理裝置的構成 7
圖2-2彩色轉灰階 9
圖2-3強化對比度示意圖 12
圖2-4黃色水平列影像灰階剖面圖之分析 13
圖2-5整張影像之灰階分析 14
圖2-6二值化示意圖 15
圖2-7二值化示意圖 16
圖2-8動態二值化示意圖 17
圖2-9動態二值化與單一閥值比較圖 18
圖3-1車道偵測系統處理流程圖 19
圖3-2車量偵測示意圖 20
圖3-3車量偵測結果 20
圖3-4車道偏移警告與前車距離系統架構圖 21
圖3-5 Peak Point之分析 22
圖3-6 Peak Point偵測結果 23
圖3-7車道線段偵測之結果 25
圖3-8(a)11x15鄰接視窗(b)小線段影像 25
圖3-9區塊群組演算法示意圖 27
圖3-10 Peak-Finding演算法流程圖 28
圖3-11以金字搭表示各運算所佔比率 29
圖3-12線段連結示意圖 29
圖3-13線段連結示意圖 32
圖3-14 中華民國車道標線示意圖 33
圖3-15車道邊界篩選的結果 34
圖3-16車道接合與選取結果 34
圖3-17減少資料量示意圖 35
圖4-1 XBase PXA255架構圖 37
圖4-2 LDWS Background示意圖 40
圖4-3 Find Peak Point 41
圖4-4 PXA255開發板實現即時車道偵測辨識系統示意圖 43
圖4-5 USB2.0即時車道偵測辨識系統 46
圖4-6 SAMSUNG610L Challenge示意圖 49
圖4-7 車道偵測辨識演法示意圖 52
圖4-8 車道偵測辨識演法示意圖 52
圖4-9 新灰階值曲線圖 55
圖4-10 Original Contrast Enhancement Flow Chart 56
圖4-11 Optimized Contrast Enhancement Flow Chart 56
圖4-12車道圖像 57
圖 4-13 Contrast Enhancement (Original vs. Optimized) 57
圖 4-14 Binary示意圖 58
圖 4-15 Binary流程圖 59
圖 4-16 Binary示意圖 59
圖 4-17車道圖像 60
圖 4-18 Binary (Original vs. Optimied) 60
圖 4-19 車道圖像 61
圖 4-20 Find Peak Point (Original vs. Optimied) 61
圖 4-21 Find Edge Point示意圖 62
圖 4-22 Find Edge Point示意圖 63
圖 4-23 Find Edge Point示意圖 64
圖 4-24 圖4-24(a)~4-24(j) 64
圖 4-25 車道圖像 67
圖 4-26 Find Edge Point (Original vs. Optimied) 67
圖 4-27 車道圖像 69
圖 4-28 Edge Connecting (Original vs. Optimied) 69

[1] Young Uk Yim and Se-Young Oh,”Three-Feature Based Automatic Lane Detection Algorithm (TFALDA) for Autonomous Driving,”IEEE Transactions on Intelligent Transportation Systems. Vol. 4, No. 4, 2003.
[2] Yuji Otsuka, Shoji Muramatsu, and Tatsuhiko Monji,“Multitype Lane Markers Recognition Using Local Edge Direction,”IEEE Intelligent Vehicle Symposium, 2002, pp1114-1119.
[3] R. M. Haralick,”Digital Step Edges from Zero Crossing of Second Directional Derivatives”, IEEE Transactions on Panem Analysis and Machine Intelligence, Vol. PAMI-6, No.1, 1984.
[4] Qing Li, Nanning Zheng, and Hong Cheng,“Springrobot: A Prototype Autonomous Vehicle and Its Algorithms for Lane Detection,”IEEE Transactions on Intelligent Transportation Systems. Vol. 5, No. 4, 2004.
[5] Jorg Kibbel, Winfried Justus, and Kay Furstenberg,“Lane Estimation and Departure Warning using Multilayer Laserscanner,”IEEE Conference on Intelligent Transportation Systems, 2005, pp777-781.
[6] Shih-Shinh Huang, Chung-Jen Chen, Pei-Yung Hsiao, and Li-Chen Fu, “On-Board Vision System for Lane Recognition and Front-Vehicle Detection to Enhance Driver's Awareness,” IEEE International Conference On Robotics And Automation, Vol. 3, 2004, pp. 2456-2461.
[7] A. L’opez, C. Ca?nero, J. Serrat, J. Saludes, F. Lumbreras, and T. Graf, “Detection of Lane Markings based on Ridgeness and RANSAC,IEEE Conference on Intelligent Transportation Systems, 2005, pp733-738.
[8] Joel C. McCall and Mohan M. Trivedi,“Video-Based Lane Estimation and Tracking for Driver Assistance: Survey, System, and Evaulation,” IEEE Transactions on Intelligent Transportation Systems, Vol. 7, No. 1, 2006.
[9] Claudio Rosito Jung and Christian Roberto Kelber, ”A Lane Departure Warning System based on a Linear-Parabolic Lane Model,”IEEE Intelligent Vehicles Symposium, 2004, pp.891 – 895.
[10] J. W. Lee, C.-D. Kee, and U. K. Yi,“A New Approach for Lane Departure Identification,”in Proc. IEEE Intelligent Vehicles Symposium, Columbus, OH, 2003, pp. 100–105.
[11] Pau-Lo Hsu I, Hsu-Yuan Cheng I, Bol-Yi Tsuei , and Wen-Jing Huang, “The Adaptive Lane-Departure Warning System,”Proceedings of the 41st SICE Annual Conference, Vol. 5, 2002, pp. 2867-2872.
[12] Claudio Rosito Jung and Christian Roberto Kelber,“A Lane Departure Warning System Using Lateral Offset with Uncalibrated Camera,”IEEE Conference on Intelligent Transportation Systems, , Sep. 2005, pp348-353.
[13] Pei-Yung Hsiao, and Chun-Wei Yeh,“A Portable Real-Time Lane Departure Warning System based on Embedded Calculating Technique,” IEEE Conference on Vehicular Technology, 2006. pp.7-04D1-5.
[14] Pei-Yung Hsiao, Chia-Hsiung Chen, Shin-Shian Chou, Le-Tien Li, and Sao-Jie Chen, “A Parameterizable Digital-Approximated 2D -83-
Gaussian Smoothing Filter for Edge Detection in Noisy Image,”IEEE Int’l Symp. on Circuits and Systems, Island of Kos, Greece, 2006, pp.3189-3192.
[15] Risack R. , Mohler N. and Enkelmann W. , "A video-based lane keeping assistant," Intelligent Vehicles Symposium, 2000. IV 2000. Proceedings of the IEEE 3-5 , 2000, pp.356-361.
[16] Gonzalez J.P. and Ozguner U. , "Lane detection using histogram-based segmentation and decision trees," Intelligent Transportation Systems, 2000. Proceedings. 2000 IEEE 1-3 , 2000, pp.346-351.
[16] Park S.T. , Yang, S.Y. and Jung, J.H. , "Real-time lane recognition by simulated annealing algorithm," Science and Technology, 2000. KORUS 2000. Proceedings. The 4th Korea-Russia International Symposium on Volume 3, 2000, pp.95-98.
[17] Zhihong Zeng and Songde Ma, "An efficient vision system for multiple car tracking," Pattern Recognition, 2002. Proceedings. 16th International Conference on Volume 2, 2002, pp.609-612.
[18]Pei-Yung Hsiao, Hsien-Chein Cheng, Chun-Wei Yeh, Shih-Shinh Huang and Li-Chen Fu, "Automobile lane detection system-on-chip integrated with mixed signal mode CMOS image sensor," Consumer Electronics, 2005 Proceedings of the Ninth International Symposium on 14-16, 2005, pp.320-323.
[19] Kaszubiak J. , Tornow M. , Kuhn R.W. , Michaelis B. and Knoeppel C. , "Real-time vehicle and lane detection with embedded hardware," Intelligent Vehicles Symposium, 2005. Proceedings. IEEE 6-8, 2005, pp.619-624.
[20] Yu-Ming Liang, Hsiao-Rong Tyan, Shyang-Lih Chang, Hong-Yuan Mark Liao, and Sei-Wang Chen, “Video Stabilization for a Camcorder Mounted on a Moving Vehicle,”IEEE Transactions on Vehicular Technology, Vol. 53, No. 6, Nov. 2004.
[21] Cipolla R, Robertson D., “3D models of architectural scenes from uncalibrated images and vanishing points,” the 10th International Conference on Image Analysis and Processing, p 824–829, 1999.
[22] David Schreiber, Bram Alefs, and Markus Clabian, “Single Camera Lane detection and Tracking,” IEEE Conference on Intelligent Transportation Systems, pp1114-1119, Sep. 2005.
[23] 陳遵立，＂ DSP-Based 之車路視覺系統之研究＂，中山大學電機工程研究所，碩士論文，2006
[24] 薛必智，＂在口袋型電腦平台上實現可攜式移動車道防偏離演算法＂，長庚大學電子工程研究所，碩士論文，2007
[25] VisualDSP++ 3.0 Getting Started Guide for Blackfin Family DSPs.2002.4
[26] Seung Gweon Geong,Chang Sup Kim,Dong Youp Lee,Sung KiHa,Dong Hwai Lee,Man Hyung Lee,and Hashimoto,H.,”Real-time lane detection for autonomous vehicle” IEEE International Symposium on Volume 3,pp1466-1471 vol.3,12-16 June 2001.
[27] Qiang Chen and Hong Wang, "A Real-time Lane Detection Algorithm Based on a Hyperbola-Pair Model," Intelligent Vehicles -85- -86- Symposium, 2006 IEEE 13-15, 2006, pp.510-515.
[28] 郭宗彥，＂以嵌入式平台實現車道偵測系統＂，長庚大學電子工程研究所，碩士論文，2008