{"title":"辅助驾驶自动车道检测方法","authors":"Maria C. Brad, A. A. Brad, M. Micea","doi":"10.1109/SACI58269.2023.10158656","DOIUrl":null,"url":null,"abstract":"This paper presents a machine learning-based method for detecting lanes on roads. The proposed approach includes several processing steps such as preprocessing of the original image frames, application of the Hough Line Transform for an initial detection of lanes, computation of the vanishing point to determine the horizon line, and region of interest (ROI) determination. Additionally, the method compensates for the unknown position of the camera facing the road by cropping a triangle-shaped perspective area. To correct errors caused by road discoloration and cracks, a color mask for white and yellow pixels is used. The orientation of the lanes is determined by analyzing the slope of the lines, and the lane coordinates are linked to the image center. The proposed method uses the U-Net neural network and the implementation is based on the Python programming language and OpenCV image processing library. In the final section we also present a comparison with a lane detection method based on convolutional neural networks and discuss the results.","PeriodicalId":339156,"journal":{"name":"2023 IEEE 17th International Symposium on Applied Computational Intelligence and Informatics (SACI)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Method for Autonomous Lane Detection in Assisted Driving\",\"authors\":\"Maria C. Brad, A. A. Brad, M. Micea\",\"doi\":\"10.1109/SACI58269.2023.10158656\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper presents a machine learning-based method for detecting lanes on roads. The proposed approach includes several processing steps such as preprocessing of the original image frames, application of the Hough Line Transform for an initial detection of lanes, computation of the vanishing point to determine the horizon line, and region of interest (ROI) determination. Additionally, the method compensates for the unknown position of the camera facing the road by cropping a triangle-shaped perspective area. To correct errors caused by road discoloration and cracks, a color mask for white and yellow pixels is used. The orientation of the lanes is determined by analyzing the slope of the lines, and the lane coordinates are linked to the image center. The proposed method uses the U-Net neural network and the implementation is based on the Python programming language and OpenCV image processing library. In the final section we also present a comparison with a lane detection method based on convolutional neural networks and discuss the results.\",\"PeriodicalId\":339156,\"journal\":{\"name\":\"2023 IEEE 17th International Symposium on Applied Computational Intelligence and Informatics (SACI)\",\"volume\":\"61 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-05-23\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2023 IEEE 17th International Symposium on Applied Computational Intelligence and Informatics (SACI)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/SACI58269.2023.10158656\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 IEEE 17th International Symposium on Applied Computational Intelligence and Informatics (SACI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SACI58269.2023.10158656","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Method for Autonomous Lane Detection in Assisted Driving
This paper presents a machine learning-based method for detecting lanes on roads. The proposed approach includes several processing steps such as preprocessing of the original image frames, application of the Hough Line Transform for an initial detection of lanes, computation of the vanishing point to determine the horizon line, and region of interest (ROI) determination. Additionally, the method compensates for the unknown position of the camera facing the road by cropping a triangle-shaped perspective area. To correct errors caused by road discoloration and cracks, a color mask for white and yellow pixels is used. The orientation of the lanes is determined by analyzing the slope of the lines, and the lane coordinates are linked to the image center. The proposed method uses the U-Net neural network and the implementation is based on the Python programming language and OpenCV image processing library. In the final section we also present a comparison with a lane detection method based on convolutional neural networks and discuss the results.