以車載光達技術輔助公路路面特徵模型建構

車載光達系統具備機動性高且可快速獲取大量三維空間資訊的優勢，已廣泛應用於工程領域。而隨著交通建設發展與都市擴張，道路特徵日益複雜且攸關用路人安全，加上近年來所興起自動駕駛系統、三維數位城市以及導航圖資更新等皆須引入路面特徵資料規畫，因此建立自動化的道路模型建構技術顯得格外重要。本研究目標為發展一套以車載光達技術為主的公路路面特徵建構流程，研究中首先利用兩階段非路面點雲濾除程序萃取路面，接著將點雲影像化並引入「分治法」演算概念，將反射強度影像自動二值化，此外配合型態影像處理、物件標記等方式將候選物件群聚分析與編組，並透過主成分分析將候選物件定向，最後再以模版匹配完成物件的辨識與萃取。數值實驗成果顯示，透過本研究所建立之自動化處理流程可快速將車載光達點雲之路面特徵物件分類萃取，各式路面特徵之分類成果品質皆可達約90%，此外透過誤差預算分析可掌握各項資料處理程序所造成錯誤分類之比例，再搭配適當改進策略後將可進一步提升分析品質，使建構之技術能更具體落實於交通與都市工程應用實務。

關鍵字

車載光達系統；路面特徵資料庫；模板匹配；分類品質指標；誤差預算分析

並列摘要

The advantage of Mobile Mapping System lies in its high mobility and efficiency in collecting massive 3D spatial information. Therefore, it has been applies in many fields of engineering frequently. As urban roads expand over time, road surface features have become more complicated and critical to driving safety. In addition, road surface features are also applied to the fields of autonomous vehicle system, 3D cyber city construction, and navigation information renewal. Therefore, constructing a road surface feature database by an automatic process is especially important. In this study, a fully automatic reconstruction of road surface features based on terrestrial mobile LiDAR technique is proposed. The complete analysis steps are developed and an experiment has been carried out using real LiDAR dataset in a case study. The results revealed that an efficient road surface reconstruction can be achieved using the proposed approach, giving a generally 90% classification accuracy. Finally, through the error budget analysis, the uncertainty associated with each analyzing stage can be explicitly assessed. As a consequence, the system configuration can be further optimized and the resulting quality of the proposed approach can be assured.

並列關鍵字

Terrestrial Mobile LiDAR System ； Road Surface Features ； Template Matching ； Feature Reconstruction ； Error Budget Analysis

參考文獻

Chen, X., Kohlmeyer, B., Stroila, M., Alwar, N., Wang, R., & Bach, J. (2009). Next generation map making: georeferenced groundlevel lidar point clouds for automatic retroreflective road feature extraction. 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, 488 - 491.

Danescu, R., & Nedevschi, S. (2010). Detection and classification of painted road objects for intersection assistance applications. International IEEE Conference on Intelligent Transportation Systems (ITSC), 433 - 438.

Fan, H., Yao, W., & Tang, L. (2014). Identifying man-made objects along urban road corridors from mobile lidar data. IEEE, Geoscience and Remote Sensing Letters, 11(5).

Han, J.Y., Guo, J., & Jiang,Y.S. (2013). Monitoring tunnel profile by means of multi-epoch dispersed 3-d lidar point clouds. Tunnelling and Underground Space Technology, 186 - 192.

Jaakkola, A., Hyyppa, J., Hyyppa, H., & Kukko, A. (2008). Retrieval algorithms for road surface modelling using laser-based mobile mapping. Sensors, 8(9), 5238 - 5249.

國際替代計量

以車載光達技術輔助公路路面特徵模型建構

全文下載

主題瀏覽