Lane Detection of Curving Road for Structural Highway With Straight-Curve Model on Vision
===
###### Huifeng Wang , Yunfei Wang , Xiangmo Zhao, Guiping Wang, He Huang, and Jiajia Zhang

## Keyword
Structural road, curve, region-of-interest, multi- nomial model, lane detection.

## Content
本文提出了一種基於直線曲線模型的彎道檢測算法，該方法在大多數彎道條件下具有良好的適用性。
首先，該方法通過分析道路圖像的基本特徵，將道路圖像分為感興趣區域和道路背景區域。
感興趣的區域進一步分為直線區域和曲線區域。同時建立了直線曲線數學模型。通過使用改進的霍夫變換獲得直線模型的數學方程。
根據道路線的連續性和直線模型與曲線模型之間的切線關係建立多項式曲線模型。然後，通過曲線擬合法求解曲線模型方程的參數。
最後，分別實現了對直線和曲線的檢測和識別，並重建了車道線，實驗表明，該方法能夠準確識別彎道線，提供有效的交通信息，進行預警，具有一定的通用性。

### 實驗方法
#### 1. 分割目標區域
在近場使用直線方程式；遠場使用圓方程式
##### 直線與曲線的車道模型
根據道路結構特點：1.具有平整度 2.道路線為實線連續或虛線連續，且直線車道線和曲線車道現已連接且相切因此可以建立直線曲線模型。
##### ROI設定
道路分佈通常包括背景環境區域和主要車道區域，透過尋找道路圖像變化設置水平線獲得
ROI，每一行二值化從頂部開始累積，搜尋相鄰兩行之間像素累積差值的最大差值找出水平線位置
#### 2. 車道模型
根據ROI特徵可以發現近視場的車道線近乎直線，遠視場的車道近似彎曲
##### 直線模型
設為y=ax+b，需求兩個未知參數。透過霍夫變換可以取得線性模型方程參數
##### 曲線
通過分析和比較圓形，雙曲線[46]，廣義曲線[47]，迴旋曲線的優缺點，可以在車道遠處使用車道線曲線的曲線模型，並建立二次曲線模型。
透過直線方程式可以求出四個參數並可以獲得曲線完整數學模型

#### 3. 預警模型
當汽車在直線道路上正常行駛時，僅使用直線模型跟踪車道線，此時，僅執行直線道路的駕駛偏離預警。 車輛進入彎道時，需要應用直線曲線模型。 同時，執行駕駛偏差和彎道預警。
##### 駕駛偏離預警
根據近車道直線區域判斷駕駛車輛的預警偏差
a、b：表示實線車道；虛線：代表車道中心點；大虛線p：表示圖像垂直線（駕駛方向）
以霍夫方程及直線方程算出λ，為直行偏差角度
##### 曲線提醒
（1）霍夫變換檢測參數直線並獲得線性方程
（2）計算左右直線車道的交點位置
（3）確定曲線消失點與直線交點的相對位置以及曲線的彎曲方向初步預測。直線交點的左側是向左彎曲的方向，直線交點的右側是向右彎曲的方向
（4）計算交點之間左右車道線上的特徵像素數點和圖像中的拐點，並比較結果以確定車道線的彎曲方向。

### 實驗結果

### 結論
通過建立直線和曲線模型並分析道路圖像的特徵，提出了一種基於直線曲線模型的車道線檢測算法。它可以更好地解決車道線的準確檢測，在實際應用中具有重要意義。實驗結果表明，該算法能夠準確識別道路車道線，並給出車輛的偏離信息和彎道方向。改善在彎道條件下的車輛的主動安全駕駛和輔助駕駛具有重要意義。

## Review
本篇使用直線方程與曲線方程的關係切分遠近視角套用不同方程式，又因兩方程式會有共同解解出未知參數將直線與曲線擬合作完，再以駕駛角度搭配攝影機垂直加上行徑方向算出夾角做出偏離預警；以消失點之於中心消失點作為彎曲頻估並提醒。

## Reference
[1] H. Li, W. Wei, and L. Sun, “Discussion on the present situation and development trend of China’s transportation,” Group Article’s Heaven Earth, vol. 11, p. 283, Feb. 2012.
[2] C.Sunetal.,“Thesafetyevaluationofcurvedrivingbasedontheanalysis of vehicle’s lateral stability,” Traffic Inf. Saf., vol. 32, no. 6, pp. 95–100, Jan. 2014.
[3] G. Yu et al., “The lateral tilt stability’s analysis and The lateral turn warn- ing’s research for vehicle’s driving of curve,” J. Beijing Univ. Technol., vol. 40, no. 4, pp. 574–579, Apr. 2014.
[4] C. Wang, “Study on road traffic safety analysis and countermeasures in curve,” Chang’an Univ., Xi’an, China, 2010.
[5] D. J. Findley et al., “Modeling the impact of spatial relationships on horizontal curve safety,” Accident Anal. Prevention, vol. 45, no. 2, pp. 296–304, Feb. 2012.
[6] G. Xiao, and Y. Chen, “Design of road corner warning system based on microwave sensor,” Transducer Microsyst. Technol., vol. 33, no. 17, pp. 94–107, Oct. 2014.
[7] C. Ma et al., “Lane detection using heuristic search methods based on color clustering,” in Proc. Int. Conf. Commun., Circuits Syst., 2010, pp. 368–372.
[8] G.Liu,S.Li,andW.Liu,“Lanedetectionalgorithmbasedonlocalfeature extraction,” in Proc. Chin. Autom. Congr., 2013, pp. 59–64.
[9] P. Wu, C. Chang, and C. Lin, “Lane-mark extraction for automobiles under complex conditions,” Pattern Recognit., vol. 47, no. 8, pp. 2756– 2767, Aug. 2014.
[10] H. Peng et al., “Lane detection algorithm based on extended Kalman filter,” J. Optoelectron. Laser, vol. 26, no. 3, pp. 567–574, Mar. 2015.
[11] M. B. Paula and C. R. Jung, “Real-time detection and classification of road lane markings,” in Proc. 26th Conf. Graph., Patterns Images, 2013, pp. 83–90.
[12] J. Huang et al., “Lane marking detection based on adaptive threshold segmentation and road classification,” in Proc. IEEE Int. Conf. Robot. Biomimetics, 2014,pp. 291–296.
[13] R. Gopalan et al., “A learning approach toward detection and track- ing of lane markings,” IEEE Trans. Int. Transp. Syst., vol. 13, no. 3, pp. 1088–1098, Mar. 2012.
[14] D.C.Hernandez,V.Hoang,andK.H.Jo,“Vanishingpointbasedimage segmentation and clustering for omni-directional image,” Intell. Comput. Theories Technol., vol. 16, no. 3, pp. 541–550, Mar. 2013.
[15] Y. Wang, X. Wang, and C. Wen, “Gradient-pair constraint for structure lane detection,” J. Image Graph., vol. 17, no. 6, pp. 657–663, Jun. 2012.
[16] F. You et al., “Lane detection algorithm for night-time digital image based on distribution feature of boundary pixels,” J. Opt. Soc. Korea, vol. 17, no. 2, pp. 188–199, Feb. 2013.
[17] H. Yoo, U. Yang, and K. Sohn, “Gradient-enhancing conversion for illumination-robust lane detection,” IEEE Trans. Intell. Transp. Syst., vol. 14, no. 3, pp. 1083–1094, Mar. 2013.
[18] J. Son, H. Yoo, S. Kim, and K. Sohn, “Real-time illumination invariant lane detection for lane departure warning system,” Expert Syst. Appl., vol. 42, no. 4, pp. 1816–1824, Apr. 2015.
[19] C.Hou,“Researchonvision-basedlanemarkingsdetectiontechnology,” Southwest Jiaotong Univ., Chengdu, China, 2014.
[20] J. Deng and Y. Han, “A real-time system of lane detection and tracking based on optimized RANSAC B-spline fitting,” in Proc. Res. Adaptive Convergent Syst., 2013, pp. 157–164.
[21] K H. Lim et al., “Lane detection and Kalman-based linear parabolic lane tracking,” in Proc. Int. Conf. Intell. Human-Mach. Syst. Cybern., 2009, pp. 351–354.
[22] A. Mammeri, A. Boukerche, and G. Lu, “Lane detection and tracking system based on the MSER algorithm, Hough transform and Kalman filter,” in Proc. ACM Int. Conf. Model., Anal. Simul. Wireless Mobile Syst., 2014, pp. 259–266.
[23] W. Li et al., “A lane marking detection and tracking algorithm based on sub regions,” in Proc. Int. Conf. Inf. Cybern. Comput. Soc. Syst., 2014, pp. 68–73.
[24] C. Chang and C. Lin, “An efficient method for lane-mark extraction in complex conditions,” in Proc. Int. Conf. Ubiquitous Intell. Comput., Int. Conf. Autonomic Trusted Comput., 2012, pp. 330–336.
[25] C.Low,H.Zamzuri,andS.A.Mazlan,“Simplerobustroadlanedetection algorithm,” in Proc. Int. Conf. Intell. Adv. Syst., 2014, pp. 1–4.
[26] T. Tan, S. Y. Yin, and L. B. Liu, “Efficient lane detection system based on monocular camera,” in Proc. IEEE Int. Conf. Consum. Electron., Mar. 2015, pp. 202–203.
[27] Q. Chen and H. Wang, “A real-time lane detection algorithm based on a hyperbola-pair model,” in Proc. Intell. Veh. Symp., Sep. 2006, pp. 510–515.
[28] W. R. Liu and S. T. Li, “An effective lane detection algorithm for structured road in urban,” in Proc. Int. Conf. Intell. Sci. Big Data Eng., Oct. 2013, pp. 759–767.
[29] L. Liang, “Lane detection algorithm based on deep learning,” Beijing Jiaotong Univ., Beijing, China, 2018.
[30] A. Mammeri, A. Boukerche, and Z. Tang, “A real-time lane marking localization, tracking and communication system,” Comput. Commun., vol. 73, no. 2, pp. 229–233, Feb. 2015.
[31] J.KimandM.Lee,“Robustlanedetectionbasedonconvolutionalneural network and random sample consensus,” in Proc. Int. Conf. Neural Inf. Process., Nov. 2014, pp. 454–461.
[32] L. Anirban and R. Vikas, “An Empirical Evaluation of various Deep Learning Architectures for Bi-Sequence Classification Tasks,” in Proc. 2016th Int. Conf. Computat. Linguist.: Tech. Papers, Osaka, Japan, Dec. 2016, pp. 2762–2773.
[33] B. Tan, “Influence of road linear design on traffic safety,” Res. The- ory Urban Construction, Electron. Ed., vol. 5, no. 24, pp. 36–41, May 2014.
[34] M.-W. K. NingYang, P. Schonfeld, and M. K. Jha, “Multi-objective highway alignment optimization incorporating preference information,” Transp. Res. C, vol. 40, pp. 36–48, Feb. 2014.
[35] L. Guo et al., “Research and application of curve detection based on road model,” Traffic Inf. Saf., vol. 30, no. 3, pp. 141–146, Mar. 2012.
[36] X.Chu,“Studyonlanedeparturewarningsystemtheoryandalgorithms based on machine vision,” Hunan Univ., Changsha, China, 2013.
[37] J.GaoandW.Wang,TheLinearDesignofHighway.Zhengzhou,China: Yellow River Water Conservancy Press, 2005.
[38] D. Ding, C. Lee, and K. Y. Lee, “An adaptive road ROI determination algorithm for lane detection,” J. Inst. Electron. Inf. Eng., vol. 51, no. 1, pp. 1–4, Jan. 2014.
[39] S. Nakamura and F. Saitoh, “Binarization of the image due to dynamic convex quadrilateral region segmentation using GA,” IEEE Trans. Elec- tron. Inform. Syst. C, vol. 134, no. 5, pp. 686–693, 2014.
[40] Z. Qu et al., “Straight-line based image registration in Hough parameter space,” in Proc. IEEE Int. Workshop Multi-Platform/Multi-Sensor Remote Sens. Mapping, Mar. 2011, pp. 1–7.
[41] K. Chishima, Y. Sakai, and K. Arakawa, “A method of scratch removal from old video film with Hough transform: Consideration for the nearly- vertical straight line element,” in Proc. IEICE Gen. Conf., Feb. 2010, pp. 68–74.
[42] R. Duan et al., “A fast detection algorithm of straight line based on improved Hough transform,” J. Instrum., vol. 31, no. 12, pp. 2774–2780, Dec. 2010.
Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on November 18,2020 at 03:38:34 UTC from IEEE Xplore. Restrictions apply.5330
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 68, NO. 6, JUNE 2019
[43] O. Khalifa O et al., “A hyperbola-pair based lane detection system for vehicle guidance,” Lecture Notes in Eng. Comput. Sci., vol. 12, no. 3, pp. 32–38, Jan. 2011.
[44] W. Liu et al., “Road detection by using a generalized Hough transform,” Remote Sens., vol. 9, no. 590, pp. 96–109, Jun. 2017.
[45] A. Singh and V. Attri, “Performance evaluation of lane detection image based on additive Hough transform algorithm,” Int. J. Adv. Res. Comput. Sci., vol. 8, no. 5, pp. 187–195, May 2017.
[46] F. Liu et al., “A lane line detection and tracking algorithm based on vision,” J. Tongji Univ., Natural Sci. Ed., vol. 38, no. 2, pp. 223–229, Feb. 2010.
[47] H. Loose, U. Franke, and C. Stiller, “Kalman particle filter for lane recognition on rural roads,” in Proc. IEEE Intell. Veh. Symp., 2009, pp. 60–65.
[48] H. Kong, J. Y. Audibert, and J. Ponce, “Vanishing point detection for road detection,” in Proc. IEEE Conf. Comput. Vision Pattern Recognit., Aug. 2009, pp. 96–103.
[49] B. Liu and T. Ceng, “Lane detection based on improved least square fitting,” Inf. Technol., vol. 4, no. 13, pp. 192–200, Apr. 2015.
[50] C. R. Jung and C. R. Kelber, “Lane following and lane departure us- ing a linear-parabolic model,” Image Vision Comput., vol. 23, no. 13, pp. 1192–1202, Jun. 2005.
[51] Y. Wang, “Research and implementation of lane line recognition algorithm based on edge,” Northeastern Univ., Boston, MA, USA, Jun. 2008.
[52] B. Yu, W. Zhang, and Z. Gong, “Lane departure warning system based
on machine vision,” J. Southeast Univ., Natural Sci. Ed., vol. 39, no. 5,
pp. 928–932, May 2008.
[53] Q. Jide, C. Bin, Q. Jiye, and C. Gang, “Fast lane detection algorithm
based on region of interest model,” J. Univ. Electronic Sci. Technol. China,
vol. 47, no. 3, pp. 356–361, Mar. 2018.
[54] X. Wang, Y. Liu, and D. Hai, “Lane detection method based on dou-
ble ROI and varied-line-spacing-scanning,” J. Command Control, vol. 3,
no. 2, pp. 154–159, Feb. 2017.
[55] U. Ozgunalp and S. Kaymak, “Lane detection by estimating and us-
ing restricted search space in Hough domain,” Procedia Comput. Sci.,
vol. 120, pp. 148–155, 2017.
[56] F.Liu,M.Tian,andZ.Hu,“Researchonvision-basedlanedetectionand
tracking for intelligent vehicle,” J. Tongji Univ., Natural Sci. Ed., vol. 35,
no. 11, pp. 1535–1541, Nov. 2007.
[57] X. Chu, K. Wang, and W. Zhang, “A study on a recognition algorithm
of curved lane based on piecewise straight line model,” Automot. Eng., vol. 39, no. 12, pp. 1141–1149, Dec. 2012.

###### tags:`Finish` `論文`