Effective and Low-cost Lane-based Map Localization for Vehicle-Centric Route Generation
2026-06-15 • Multimedia
MultimediaComputer Vision and Pattern Recognition
AI summaryⓘ
The authors developed OLRA, a system that helps cars understand the best driving route by combining map information with real-time camera images of lane markings. This approach improves how well the car knows its location and keeps the suggested route clear from the driver's viewpoint. They created new ways to measure how good such systems are and tested OLRA against an existing method called OpenPilot. Their results showed OLRA works better on tricky roads and for routes farther than 20 meters. This work could help make route-guidance systems cheaper and more accurate.
driver-centric route representationmap localizationlane markingsvehicle localizationnavigation routesEuclidean errorOpenPilotnuScenes datasetroute evaluation metricsvisual route consistency
Authors
Hong-Shiang Lin, Jung-Hsin Chen, Yu-Luen Tzeng, Wei-Hao Chen, Yi-Chen Lee, Li-Jhe Chen, Peng-Yuan Chen
Abstract
Driver-centric route representation plays a vital role in intuitive driving guidance systems. This paper presents OLRA, a low-cost, map-localization-based framework that derives driver-view-aligned routes by matching map-based navigation routes with camera-detected lane markings. This alignment process mutually enhances vehicle localization accuracy and visual route consistency. To bridge the evaluation gap across different paradigms, we introduce practical route evaluation metrics and benchmark OLRA against OpenPilot, a representative direct-generation approach. Experimental results on the nuScenes dataset demonstrate that OLRA outperforms OpenPilot in complex road segments and in route estimation at distance beyond 20 meters, achieving lower overall Euclidean error. This study is expected to promote future research in low-cost, maplocalization-based route generation methods.