Automatic Generation of 3D Building Models from Point Clouds - Geoinformatics for Intelligent Transportation - page 113

Fig. 3 Datasets: #1 — Czech Land Survey Office [left] and #2 — Krkono š e Mountains National
Park [right]. Top Raw point clouds (colour hypsometry in relation to the height of the terrain).
Bottom Classified point clouds: terrain (gray), buildings (red), other (yellow), unprocessed (cyan)

dataset N o Table 1 Comparison of input data; #

—

# Owner of data Area Average density

(p/m 2 )

1 Czech Land Survey Of ce Rokytnice nad Jizerou

(Czech Republic)

1.2

2 Krkono e Mountains

National Park

Rokytnice nad Jizerou
(Czech Republic)

8.2

š

3 ENVI LiDAR sample data Kleinwolkersdorf (Austria) 21.8

geometric primitives by interleaving planes through the point cloud. The software
interleaves planes exclusively from the points that were classi ed into the class

buildings. This approach reduces the amount of data that must be analyzed (Fig. 5 ).