LAS is the binary ASPRS industry standard for point clouds that has underpinned LiDAR data exchange for two decades: each point carries XYZ, intensity, and a classification code (ground, vegetation, buildings). An honest note about our export: we do NOT capture LiDAR from an aircraft — a real point cloud comes from a sensor, not from OpenStreetMap tags. Our exporter builds a structural cloud from OSM vector geometry: a building centroid plus a roof point at height get class 6 (Building), points along roads get class 2 (Ground), and POIs get class 1. The output is a valid LAS 1.2 (Point Data Record Format 0) that opens in CloudCompare, QGIS, and PDAL, but it represents schematic OSM points rather than a dense surface scan.
主要特点

关于 LAS 格式
LAS is the ASPRS standard for point clouds that laser scanning data exchange rests on: every point carries coordinates, intensity and a classification code — ground, vegetation, building. The format is binary, compact and read by every point cloud processing tool.
The main thing first: a real point cloud is captured by a laser sensor, not by the OpenStreetMap database. Our export builds a structural cloud out of vector geometry: a building contributes a base point and a roof point at its height, points are spread along roads, and points of interest become separate marks. Geometrically this is valid LAS, but in content it is a schematic of a city rather than a survey of the surface.
Hence the scenarios: test the import in your own pipeline without hunting down a real dataset; show the structure of the file and the ASPRS class system; get predictable data for debugging filtering scripts.
Technically this is LAS version 1.2 with point records of format zero — twenty bytes per point: coordinates, intensity, return number, class, scan angle. Buildings get class 6, roads class 2 (ground), points of interest class 1. The header is correct, so the file is read without errors by laspy, PDAL, CloudCompare and QGIS. The limit is 500,000 points per export.
打开 LAS 文件的程序
以下程序可以打开、编辑和处理从 OSM2CDR 导出的 LAS — ASPRS Point Cloud (LiDAR format)(.las)文件:
格式规格
- 开放标准
- 支持 3D 坐标
- 无损(无质量损失)
- 存储要素属性
- 输出:1 个文件
- A LAS 1.2 (PDRF 0) structural cloud synthesized from OSM: buildings -> class 6, roads -> class 2, POIs -> class 1, capped at 500,000 points.
谁在使用 LAS 地图
如何导出 LAS
导出后如何打开
- CloudCompare (free) — drag the .las into the window; on loading, keep the offset it suggests, otherwise the points will drift far from the center of the scene.
- PDAL — pdal info map.las shows the header and statistics, and a filter on Classification[6:6] leaves only the buildings.
- QGIS (free) — Layer → Add Point Cloud Layer; the points can be colored by class or by height.
- Python — laspy.read("map.las") gives arrays of coordinates and classes for your own computations.
- Comparing with real LiDAR — download a tile from an open portal (USGS 3DEP, AHN, swissSURFACE3D) and lay our cloud over it as a reference against the buildings.
LAS 格式的热门地图
导出 LAS — ASPRS Point Cloud (LiDAR format) 地图
从 OpenStreetMap 下载专业的 LAS 地图。即时生成,任意区域。