利用資料蒐評法自動萃取空載光達資料中的屋頂面資訊

Automatic Extraction of Building Roofs by Data Snooping from Airborne LIDAR Data

空載光達資料是大量分布於地表的三維點位資料，經過處理之後，可以應用於高精度的DEM生產；萃取如建物、道路或者樹木等地物，作為景觀規劃或是其他方面之應用；甚至可以用來進行災害評估。由於空載光達資料中隱含著空間中有意義的點、線、面特徵，對於萃取其特徵建構都市區或含建物區的建物模型供後續應用是目前相當重要的研究主題。而由包含大量精確平面特徵的光達資料中，進一步處理或結構化之後萃取有意義的面特徵，甚至是更有意義的地物資訊供後續應用則需要發展不同的演算法。本文即嘗試由包含建物資訊區域範圍內的光達資料中，萃取屬於建物屋頂面資訊供後續處理應用；也就是在假設地表建物屋頂面是由三維平面所建構而成的前提之下，利用資料蒐評法發展可以確實排除不屬於屋頂面上光達資料的演算法，並精確決定建物屋頂面的資訊供建立建物模型之用。 LIDAR data are consisted of a large number of 3-D points on the terrain surface. After processing, LIDAR data can be used for the generation of high accuracy DEM, for the extraction of terrain object, e.g. buildings, roads and trees, as well as for disaster assessments. Urban LIDAR data contains lots of meaningful 3-D planes for building reconstruction. Therefore many algorithms are developed to extract the meaningful 3-D plane features. Based on data snooping theory, this paper proposed an algorithm to exclude non-roof point data and extract building roof planes from LIDAR data for the application of building reconstruction. The test results show the feasibility of the proposed algorithm.

LIDAR data are consisted of a large number of 3-D points on the terrain surface. After processing, LIDAR data can be used for the generation of high accuracy DEM, for the extraction of terrain object, e.g. buildings, roads and trees, as well as for disaster assessments. Urban LIDAR data contains lots of meaningful 3-D planes for building reconstruction. Therefore many algorithms are developed to extract the meaningful 3-D plane features. Based on data snooping theory, this paper proposed an algorithm to exclude non-roof point data and extract building roof planes from LIDAR data for the application of building reconstruction. The test results show the feasibility of the proposed algorithm.