以物件式分類與車載光達資料辨識工址標誌

Using Mobile LiDAR Data with Objective Classification to Identify the Site Signs

李佩玲(P. L. Li) ； 余騰鐸(T. T. Yu) ； 莊鈞凱(C. K. Chuang) ； 鄭詠心(Y. S. Chen)

《鑛冶：中國鑛冶工程學會會刊》 58卷1期 (2014/03) Pp. 70-82

https://doi.org/10.30069/MM.201403_58(1).0006

摘要

任何工址、礦區或道路均有特定目的之標誌，本研究針對車載光達資料對於標誌以物件式分類法則進行辨識。對此研究中以高程變化大且視野狹窄之山區道路為研究區域，並於最後加入平坦公路路段辨識結果進行探討。其結果顯示，透過本法探測研究區內山區道路59個道路標誌，其結果可達84.7%準確率；而應用於研究區內平坦公路路段81個道路標誌，亦也有81.4%準確率，相關使用限制及說明將於文中進行探討，未來亦可供公路總局相關單位進行道路標誌建置及清查時參考。

關鍵字

車載光達 ； 道路標誌 ； 物件式分類

並列摘要

There are specific signs of each site and mining area, and the target of this research is finding signs with objective classification. In order to reach such goal, this study first focus on identifying road signs with mobile LiDAR data in the mountain roads, which has severely elevation change and poor line of sight. Eventually, we add in the identified results in flat highway areas, to serve as complement to discuss the method applicability in different areas. In this study, total 59 road signs in mountain area can be correctly identified and the accuracy is 84.7%; 81 road signs in flat highway areas can be correctly identified and the accuracy is 81.4%. On the other hand, the restriction also would be discussed in this article and could offer the highways authority to build road signs or inventory check.

並列關鍵字

Mobile LiDAR ； Road Signs ； Objective Classificatio