Semi-automatic Use of High Resolution Images and Digital Elevation Models for Counting and Identification of Forest Trees

This paper presents a semi-automatic and systematic computational approach intending to count and localize different species of trees in zones of dense forest. Comparative analysis of the application of multi-spectral high resolution images and aerial LIDAR data are presented. A method of image segmentation (object-oriented) and a combination of the digital elevation model (DEM) of the crown trees with the reflectance values of the three visible (RGB) and near infrared (NIR) bands of the aerial image are used. In this field of study, Hyypa et al (2004) described different existing algorithms and Suarez et al. (2005) presented an efficient methodology to implement a forest inventory by fusing of aerial LIDAR data and digitized aerial photos.


Presented at:
Geocomputation 2007, NUI Maynooth, Ireland, September 3-5, 2007
Year:
2007
Keywords:
Laboratories:




 Record created 2007-09-28, last modified 2018-03-13


Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)