Consultants for Surveying Engineering
Welcome to AVT,
Your partner for surveying

Classification

Automatic methods rely on neighborhood relationships between points (e.g. groups of points defining a common plane), or properties of single points (e.g. echo number or intensity in case of laser scanning, or color in case of images). Manual improvement involves profiles and hillshade visualizations. 

The quality of automatic classification may reach 75%, raising it to beyond 90% often requires lots of manual work.