Lane Detection for Autonomous Vehicles

Abstract

Nowadays,autonomousvehiclesaredevelopingrapidly toward facilitating human car driving. One of the mainissuesisroadlanedetectionforasuitableguidancedirectionand car accident prevention. This paper aims to improve andoptimize road line detection based on a combination of cameracalibration,theHoughtransform,andCannyedgedetection.Thevideo processing is implemented using the Open CV library withthe novelty of having a scale able region masking. The aim of thestudy is to introduce automatic road lane detection techniqueswiththeuser’sminimummanualintervention.

Introduction

This study addresses the important issue of road safety by improving road lane detection techniques using image processing. Accurate lane detection helps reduce accidents caused by human error. Traditional image processing methods, such as combining Canny edge detection and Hough transform, are effective, cost-efficient, and relatively simple to implement compared to machine learning approaches.

The proposed framework enhances existing methods through camera calibration (to correct lens distortion), optimized grayscale conversion, noise removal via an adaptive Gaussian filter, and automatic masking of the region of interest to focus only on the relevant road area in images. These improvements reduce false detections and increase accuracy.

Dynamic lane detection is performed on videos and images captured from cameras mounted on cars, including real-world dangerous road scenarios. The process involves:

• Camera calibration to correct lens distortion.

• Grayscale conversion to simplify image data.

• Noise removal with an optimized Gaussian blur filter.

• Automatic masking to limit analysis to road areas.

• Canny edge detection to identify edges.

• Probabilistic Hough transform to detect lane lines from edges.

• Calculation of line slopes to distinguish left and right lanes.

The system minimizes manual intervention and works in real-time or offline on microchips or computers, providing timely alerts to drivers if lane crossing or obstacles are detected. Experimental results on test videos and images demonstrate improved lane detection performance.

Conclusion

In this study, our aim was to propose and develop a pureimageandvideoprocessingapproachtodetectthelanesontheroadautomaticallyinthemostpossibleoptimizedwayfor better usage in real-time processing. The main method ofthis study was based on previous works on Hough transform-based road lane detection and we added an automatic cameracalibration and furthermore an optimization for making andchoosing the best kernel size in Gaussian blur to make themethodscaleableforallsizesofcamera’sframesforeaseofuse.Themethodproposedbythisworkstartedwithsome prepossessing steps like camera calibration, grey scale,Gaussian blue and further, we proposed a novel idea to makethe masking the region of interest in the process completelyautomatic. As shown in the aforementioned figures and tables,the results of lane detection in this study on test images andtestvideoswereincurablyamazingandalmostfast,compatiblefor aiming the goal mentioned before as real-time processing.However, more improvements on better accuracy of the curvedetectioncanbeprovided.

References

[1] C.Y. Kuo, Y.R. Lu and S.M. Yang, On the Image Sensor Processing forLane Detection and Control in Vehicle Lane Keeping Systems, Interna-tional Program on Energy Engineering, National Cheng Kung Univer-sity, Tainan 70101, Taiwan, 8 April 2019, https://www.mdpi.com/1424-8220/19/7/1665/pdf. [2] ArchitRastogi, Computer Vision: Lane Finding Through Image Process-ing, 6 Sep 2020, https://medium.com/swlh/computer-vision-lane-finding-through-image-processing-516797e59714 [3] Qi, Wang Li, Fu Zhenzhong, Liu. (2010). Review on Camera Calibration.3354-3358.10.1109/CCDC.2010.5498574. [4] Kanan,ChristopherCottrell,Garrison.(2012).Color-to-Grayscale:Does the Method Matter in Image Recognition. PloS one. 7. e29740.10.1371/journal.pone.0029740. [5] “LearntheLatestTechSkills;AdvanceYourCareer.”Udacity,https://www.udacity.com/. [6] Chin-Pan Huang, Chaur-Heh Hsieh, Jin-Long Li. A Robust Lane De-tectionMethodUsingAdaptiveRoadMask.DepartmentofComputer and CommunicationEngineering,MingChuanUniversity,Taiwan.(2015)https://www.academia.edu/10351901/A?Robust?Lane?Detection [7] Samuel,MovehMohamad,Maziahmadsaad,ShaharilHussein,Mohamed. (2018). Development of Edge-Based Lane Detection Algo-rithm using Image Processing. JOIV: International Journal on InformaticsVisualization.2.19.10.30630/joiv.2.1.101. [8] Canny J. F., 1986. A Computational Approach to Edge Detection. IEEETrans.PatternAnal.MachineIntel,vol.PAMI-8,no.6,pp.679-697. [9] Matas, JiriGalambos, Charles. (1998). Progressive Probabilistic HoughTransform. [10] Fokkinga, Maarten. (2011). The Hough transform. J. Funct. Program..21.129-133.10.1017/S0956796810000341. [11] N. Kiryati, Y. Eldar, A.M. Bruckstein, A probabilistic Hough trans-form,PatternRecognition,Volume24,Issue4,1991,Pages303-316,ISSN0031-3203,https://doi.org/10.1016/0031-3203(91)90073-E.(https://www.sciencedirect.com/science/article/pii/003132039190073E) [12] Statistical Bulletin: School Leavers 1992/93, https://books.google.co.uk/books?id=98qnxgEACAAJ, 1995 ,DepartmentofEducation [13] R. O. Duda, and P. E. Hart, ”Use of the Hough Transformation to DetectLines and Curves in Pictures”, the Comm. ACM, Vol.15, No.1, 1972, pp.11-15. [14] Kim, SunghoonPark, Jeong-HoCho, SeongPark, SoonyoungLee,Kisung Choi, Kyoungho. (2007). Robust Lane Detection for Video-BasedNavigationSystems.2.535-538.10.1109/ICTAI.2007.20. [15] Zuroski,Kathyrn.(2021).FindingtheslopeofalineusingInteractivate. [16] Lim, Kai LiBraunl, Thomas. (2019). A Methodological Review of Vi-sualRoadRecognitionProceduresforAutonomousDrivingApplications. [17] Sharma, AnmolKumar, ManeeshGupta, RajkamalKumar, Rajesh.(2021).LaneDetectionusingPython. [18] Pansambal, SuvarnaKumar, Udaya. (2019). Lane Datasets for LaneDetection.0792-0796.10.1109/ICCSP.2019.8698065. [19] Lu,PingpingCui,ChenXu,ShaobingPeng,HueiWang,Fan.(2021).SUPER:ANovelLaneDetectionSystem.IEEETransactionsonIntelligentVehicles.PP.1-1.10.1109/TIV.2021.3071593. [20] Mongkonyong,PeerawatNuthong,ChaiwatSiddhichai,SupakornYamakita,Masaki.(2018).LanedetectionusingRandomizedHoughTransform. IOP Conference Series: Materials Science and Engineering.297.012050.10.1088/1757-899X/297/1/012050. [21] Dai,XingjianXie,JinQian,JianjunYang,Jian.(2020).LaneDetectionCombiningDetailsandIntegrity:anAdvancedMethodforLaneDetection.1-6.10.1145/3446132.3446145. [22] Fan,C.Xu,J.Di,S.Shi,X..(2013).LanedetectionbasedonparallelHoughtransform.JournalofComputationalInformationSystems.9.4893-4900.10.12733/jcis6333. [23] Xiao,F.,Jin,L.Haopeng,W.(2010).Astudyofimageretrievalbased on hough transform, 3rd IEEE Computer Science and InformationTechnology(ICCSIT),Chengdu,China. [24] Cao J, Song C, Song S, Xiao F, Peng S. Lane Detection Algorithm forIntelligent Vehicles in Complex Road Conditions and Dynamic Environ-ments. Sensors (Basel). 2019 Jul 18. SongSong Cao, J.; Song, C.; Song,S.;Xiao,F.;Peng,S.LaneDetectionAlgorithmforIntelligentVehiclesin Complex Road Conditions and Dynamic Environments. Sensors 2019,19,3166.https://doi.org/10.3390/s19143166.

Copyright

Copyright © 2025 Ayushmaan Kamboj , Dr. Archana Kumar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.