Separating the drivable and non- drivable areas on semi-structured and unstructured roads is an important task for autonomous vehicles to safely and avoid obstacles. Semi structured and unstructured roads have different intensities, normal vector angles, and curvature information than the background, and this paves the way for the design and development of an efficient detection system for drivable areas on this roads. In this paper, an effective method for detecting drivable areas is proposed that is based on important indicators of an experimental vehicles. This method calculate the information gain of features is calculated firstly to determine the sequence of feature processing. On the basis of this sequence calculate the maximum inter-class variance of features, and combined with the specific indicators of the experimental vehicle to realize the detection of drivable areas. Finally, the performance of the method is evaluated in terms of average precision, recall, and detection accuracy, and compared with the performance of existing road detection methods, including the K-nearest-neighbors classifier and the random forest classifier methods. The experimental results show that the average precision, recall, and detection accuracy of the system are 96.19%, 96.89%, and 96.72%, respectively. The method proposed here can effectively identify and classify drivable areas on semi structured and unstructured roads.
AlamASinghLJafferyZA, et al. Distance-based confidence generation and aggregation of classifier for unstructured road detection. J King Saud Univ - Comput Inf Sci2022; 34: 8727–8738.
2.
ShiXHeWHanL. The road-edge detection algorithm that adopted Hough transform. Trans Intell Syst2012; 7: 81–85.
3.
LiuZYuSZhengN. A co-point mapping-based approach to drivable area detection for self-driving cars. Engineering2018; 4: 479–490.
ChenSLiuZ. The image processing method that eliminates the interference of marking lines when edge detection recognizes road boundaries. J Beijing Univ of Technol2021; 47: 1330–1337.
6.
ZhangW. Lidar-based road and road-edge detection. In: 2010 IEEE intelligent vehicles symposium, 2010, pp.845–848. New York, NY: IEEE.
7.
ZhuXGaoMShangnianL. A real-time road-edge detection algorithm for smart cars. Beijing Union Univ2015; 29: 1–7.
8.
DuanJLongjieLKaihuaZ. Detection of the drivable area of the road ahead based on vehicle-mounted 4-line lidar. Automot Technol2016; 47: 55–62.
9.
ZouBTanLHouX. LiDAR-based detection of driveable areas of the road. J Wuhan Univ Technol2017; 41: 203–207.
10.
KongDSunLWangJ, et al. Road edge recognition algorithm based on 3D lidar point cloud. Guangxi Univ2017; 42: 855–863.
11.
KongDWangXSunL, et al. Structured road boundary extraction method for vehicle-mounted radar point clouds. J Henan Univ Sci Technol2018; 39: 57–62.
12.
LiQDuY. Research and implementation of real-time detection algorithm based on 3D LiDAR road-edge. Comput Appl Softw2017; 34: 219–222.
13.
RatoDSantosV. LIDAR based detection of road boundaries using the density of accumulated point clouds and their gradients. Robot Auton Syst2021; 138: 714–726.
14.
ZhaoJPanWWeiS, et al. Semantic segmentation of indoor point clouds based on machine learning–door as the example. Beijing Surv Mapp2018; 32: 255–259.
15.
LetardMCollinACorpettiT, et al. Classification of land-water continuum habitats using exclusively airborne topobathymetric LiDAR green waveforms and infrared intensity point clouds. Remote Sens2022; 14: 341.
16.
DuWBeltrameG. Real-time simultaneous localization and mapping with LiDAR intensity. arXiv preprint arXiv:09257, 2023.
17.
ChengXHuXTanK, et al. Automatic detection of shield tunnel leakages based on terrestrial mobile LiDAR intensity images using deep learning. IEEE Access2021; 9: 55300–55310.
18.
SchrumPTJamesonCDTateosianLG, et al. Curvature weighted decimation: a novel, curvature-based approach to improved lidar point decimation of terrain surfaces. Geomat2023; 3: 266–289.
19.
PaivaPVCogimaCKDezen-KempterE, et al. Historical building point cloud segmentation combining hierarchical watershed transform and curvature analysis. Pattern Recognit Lett2020; 135: 114–121.
20.
LaoJWangCNieS, et al. A new denoising method for photon-counting LiDAR data with different surface types and observation conditions. Int J Digit Earth2023; 16: 1551–1567.
21.
HanYYuanJSiJ, et al. Research on real-time small obstacle detection of 16-line radar point cloud. Laser Optoelectron Prog2021; 58: 487–498.
SanchezJDenisFCoeurjollyD, et al. Robust normal vector estimation in 3D point clouds through iterative principal component analysis. ISPRS J Photogramm Remote Sens2020; 163: 18–35.
24.
WangMYiF. Point cloud smoothing and denoising algorithm based on local domain change factor. AOPC 2021: optical sensing and imaging technology2021; 12065: 535–540.
25.
HashimotoTSaitoM. Normal estimation for accurate 3D mesh reconstruction with point cloud model incorporating spatial structure. In: CVPR workshops, 2019.
26.
ZhouJJinWWangM, et al. Fast and accurate normal estimation for point clouds via patch stitching. Comput Aided Des2022; 142: 121–127.
27.
YaoZZhaoQLiX, et al. Point cloud registration algorithm based on curvature feature similarity. Measurement2021; 177: 274–285.
28.
YangRYouYWuH. A method for calculating the curvature of mixed surface scattered point clouds. Surv Mapp2014; 39: 129–131.
29.
FoorginejadAKhaliliK. Umbrella curvature: a new curvature estimation method for point clouds. Procedia Technol2014; 12: 347–352.
30.
XiaoGMeiHWangQ, et al. Point cloud semantic segmentation network based on adaptive convolution and attention mechanism. In: 2022 4th international conference on intelligent control, measurement and signal processing (ICMSP), 2022, pp.939–943. New York, NY: IEEE.
31.
LiuYXuHSuX, et al. Curvature feature extraction based ICP points cloud registration method. Optoelectronic imaging and multimedia technology V2018; 10817: 31–38.
32.
LiBLiJChenG, et al. De-snowing LiDAR point clouds with intensity and spatial-temporal features. In: 2022 international conference on robotics and automation (ICRA), 2022, pp.2359–2365. New York, NY: IEEE.
33.
HanJYPerngNHLinYT. Feature conjugation for intensity-coded LIDAR point clouds. J Surv Eng2013; 139: 135–142.
34.
ReymannCLacroixS. Improving LiDAR point cloud classification using intensities and multiple echoes. In: 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS), 2015, pp.5122–5128. New York, NY: IEEE.
35.
VacekPJasekOZimmermannK, et al. Learning to predict lidar intensities. IEEE Trans Intell Transp Syst2022; 23: 3556–3564.
36.
DynNHormannKKimS, et al. Optimizing 3D triangulations using discrete curvature analysis. Math Methods Curves2001; 1: 135–146.
37.
EvansJSHudakATSensingR. A multiscale curvature algorithm for classifying discrete return LiDAR in forested environments. IEEE Trans Geosci Remote Sens2007; 45: 1029–1038.
38.
SequeiraGAfrajSSurveM, et al. Lidar point cloud analysis for vehicle contour estimation using polynomial approximation and curvature breakdown. In: 2020 IEEE 92nd vehicular technology conference (VTC2020-Fall), 2020, pp.1–6. New York, NY: IEEE.
39.
KimHWillersJKimS. Digital elevation modeling via curvature interpolation for LiDAR data. Electron J Differ Equ2016; 23: 47–57.
40.
SofiaGPirottiFTarolliP. Variations in multiscale curvature distribution and signatures of LiDAR DTM errors. Earth Surf Process2013; 38: 1116–1134.
41.
RutzingerMHöfleBKringerK. Accuracy of automatically extracted geomorphological breaklines from airborne LiDAR curvature images. Geogr Ann A Phys Geogr2012; 94: 33–42.
42.
LeiBSunZ. Visual odometry based on point cloud curvature. J Res Sci Eng2020; 2: 43–48.
43.
FritschJKuhnlTGeigerA. A new performance measure and evaluation benchmark for road detection algorithms. In: 16th international IEEE conference on intelligent transportation systems (ITSC 2013), 2013, pp.1693–1700. New York, NY: IEEE.
