The LiDAR dataset we will use is a 2006 collection, which was classified into ground, tall vegetation (> 20 feet), buildings, and so on. One characteristic of the analysis that follows is that we assume the classification to be correct, and so we are not revisiting the quality of the classification or attempting to improve it within PostGIS.
A characteristic of the LiDAR dataset is that a sample point exists for relatively flat surfaces at approximately no fewer than 1 for every 5 feet. This will inform you about how we manipulate the data.
First, let's load our dataset using the following command:
shp2pgsql -s 3734 -d -i -I -W LATIN1 -g the_geom lidar_buildings chp04.lidar_buildings | psql -U me -d postgis_cookbook