Investigation on Precision Evaluation Method of Self-Constructing Map for Autonomous Driving

自動運転用地図の自動生成における地図精度判定手法の検討

This paper proposes a method to evaluate whether constructed map contains enough amount and precision of landmark information for autonomous driving. The proposed method uses only the map information itself to judge the capability of autonomous driving and therefore the method does not need any trials of autonomous driving control nor simulation, which leads to reduction of time and cost for map evaluation. Firstly, this paper generates the map for autonomous driving using lane-marker positions detected by on-board cameras. Secondly, the generated maps are separated with a certain length of sections and the proposed map evaluation method is applied at every section of the map. The method uses statistical analytics algorithm based on Simple Linear Regression Model, which determines whether the amount and deviation of landmarks (lane-marker positions) assures lateral distance position error of less than 0.5 m. Effectiveness of the proposed method is verified using manual driving data on a test course.著者らは、自動運転に必要な地図情報を、車載カメラの認識情報のみを用いて自動的かつ低コストに生成する研究を行っている。本講演では、自動的に生成した地図情報が十分な精度を持っているか（自動運転可能なレベルであるか）を判定する地図精度判定手法を検討し、実車走行データを用いた検証結果を報告する。

• Record URL:
  https://doi.org/10.11351/jsaeronbun.49.1074
• Record URL:
  http://scholar.google.com/scholar_lookup?title=Investigation%20on%20Precision%20Evaluation%20Method%20of%20Self-Constructing%20Map%20for%20Autonomous%20Driving&author=K.%20Maeda&author=J.%20Takahashi&author=T.%20Kojima&author=P.%20Raksincharoensak&publication_year=2018
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/issn/02878321
• Authors:
  • Maeda, Kenta
  • Takahashi, Junya
  • Kojima, Takao
  • Raksincharoensak, Pongsathorn
• Publication Date: 2018-9

Language

• English
• Japanese

Media Info

• Media Type: Digital/other
• Features: Figures; References;
• Pagination: pp 1074-1079
• Serial:
  • Transactions of Society of Automotive Engineers of Japan
  • Volume: 49
  • Issue Number: 5
  • Publisher: Society of Automotive Engineers of Japan
  • ISSN: 0287-8321
  • EISSN: 1883-0811
  • Serial URL: https://www.jstage.jst.go.jp/browse/jsaeronbun
• Publication flags:

  Open Access (libre)

Subject/Index Terms

• TRT Terms: Advanced vehicle control systems; Digital mapping; Evaluation; Intelligent vehicles; Lane occupancy; Lateral placement; Precision; Statistical analysis
• Uncontrolled Terms: Landmarks
• Subject Areas: Data and Information Technology; Highways; Vehicles and Equipment;

Filing Info

• Accession Number: 01686073
• Record Type: Publication
• Source Agency: Japan Science and Technology Agency (JST)
• Files: TRIS, JSTAGE
• Created Date: Nov 20 2018 10:24AM