近年來，全球汽車技術逐漸成熟，許多國家對自動駕駛汽車進行了現場測試。除了在駕駛期間影響車輛的安全性和舒適性的障礙物之外，道路上的坑洞還可能對車輛造成危險。為了減少由不良路況引起的事故，必須準確地檢測道路邊界並檢測道路坑洞或障礙物。因此，在本研究中，應用LiDAR來檢測道路的邊界和坑洞或障礙物。假設道路車道與道路邊界之間存在台階。用於檢測道路邊界的概念是估計激光雷達與光線照射的道路上的點之間的距離。比較LiDAR與道路上兩個連續測量點之間的距離，如果這兩個點距離的差值大於閾值，則這兩個點不在同一水平面上。用於確定路面的兩個點是否位於相同水平面上的方法稱為斷點檢測算法。迭代終點擬合算法適用於適合道路輪廓。在這項研究中，斷點檢測算法擴展到包括找到道路的坑洞或障礙物。實驗結果表明，它可以成功地檢測道路邊界和坑洞。當凹坑孔的寬度大於10厘米時，誤差1.5厘米。

In recent years, global auto technology has gradually matured, and many countries have conducted on-site tests on autonomous vehicles. In addition to obstacles affecting the safety and comfort of the vehicle during driving, the pit hole on the road can also cause dangerous to the vehicle. In order to reduce accidents caused by poor road conditions, it is necessary to accurately detect road boundaries and detect road pit holes or obstacles. Therefore, in this research, LiDAR is applied to detect the boundary and pit hole or obstacles of a road. There is an assumption that there is step existed between the road lane and road boundary. The concept used for detection the boundary of a road is to estimate the distance between the LiDAR and the point on the road the light strike on. Compare the distance between the LiDAR and two continuous strike points on the road, if the difference of these two distance is large than a threshold, then these two points are not located on the same horizontal level. The method used to determine if the two stroked points of the road surface are located on the same horizontal level is called break point detection algorithm. The Iterative End Point Fit algorithm is applied to Fit road contours. In this study, the breakpoint detection algorithm is extended to including find the pit holes or obstacles of the road. Experimental results show it can successfully detect road boundaries and pit hole. When the width of the pit hole are more than 10 cm, the error is less than 1.5 cm respectively.

參考文獻
[1] BORGES, Geovany Araujo; ALDON, Marie-José. Line extraction in 2D range images for mobile robotics. Journal of intelligent and Robotic Systems, 2004, 40.3: 267-297.
[2] HAN, J., et al. Road boundary detection and tracking for structured and unstructured roads using a 2D lidar sensor. International Journal of Automotive Technology, 2014, 15.4: 611-623.
[3] WIJESOMA, Wijerupage Sardha; KODAGODA, KR Sarath; BALASURIYA, Arjuna P. Road-boundary detection and tracking using ladar sensing. IEEE Transactions on robotics and automation, 2004, 20.3: 456-464.
[4] DEMIR, S. Ö., et al. An adaptive approach for road boundary detection using 2D LIDAR sensor. In: 2017 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI). IEEE, 2017. p. 206-211.
[5] PETERSON, Kevin; ZIGLAR, Jason; RYBSKI, Paul E. Fast feature detection and stochastic parameter estimation of road shape using multiple lidar. In: 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2008. p. 612-619.
[6] CREMEAN, Lars B.; MURRAY, Richard M. Model-based estimation of off-highway road geometry using single-axis ladar and inertial sensing. In: Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006.IEEE, 2006. p. 1661-1666.
[7] KANG, Yeonsik, et al. A lidar-based decision-making method for road boundary detection using multiple kalman filters. IEEE Transactions on Industrial Electronics, 2012, 59.11: 4360-4368.
[8] BUJA, A., et al. Duda, RO & Hart, PE (1973), Pattern classification and scene analysis, Wiley, New York. Friedman, J.(1994), Flexible metric nearest neighbour classification, Technical report, Stan-ford University.
[9] YALCIN, O., et al. Detection of road boundaries and obstacles using LIDAR. In: 2014 6th Computer Science and Electronic Engineering Conference (CEEC). IEEE, 2014. p. 6-10.
[10] HAN, Jaehyun, et al. Enhanced road boundary and obstacle detection using a downward-looking LIDAR sensor. IEEE Transactions on Vehicular Technology, 2012, 61.3: 971-985.
[11] KANG, Yeonsik, et al. A lidar-based decision-making method for road boundary detection using multiple kalman filters. IEEE Transactions on Industrial Electronics, 2012, 59.11: 4360-4368.