OBJECT ORIENTED TRANSMISSION LINE CORRIDORS CLASSIFICATION USING LIDAR TECHNOLOGY AND A NON-METRIC DIGITAL CAMERA

Abstract

The increasing availability of high-resolution imagery as well as high density and
accurate Digital Surface Models (DSMs) and Digital Terrain Models (DTMs) as
provided by LIDAR (Light Detection And Ranging) technology, has been
extending the use of remote sensing data in applications that demand a higher work
scale, as transmission line design. The result of the processed LIDAR data consists
in three-dimensional information about cables, structures as well as all obstacles
along the corridor in a form of a point cloud with X,Y,Z coordinates and intensity
value. Post-processing procedures are needed in order to discriminate features and
to derive additional information. In order to discriminate the obstacles along the line
corridor, information derived from LIDAR data as obstacles height and the intensity
image, as well as color imagery from a non-metric digital camera are used. The
integrated use of LIDAR data and color aerial photography provides more accurate
classification result as well as the discrimination of additional features. The
classification method used is based on an object-oriented analysis, which is
considered the most adequate procedure when working with high-resolution
imagery. In this work, the objects are generated using multiresolution segmentation,
FNEA (Fractal Net Evolution Approach), hierarchical network of image objects,
which represents image information in different spatial resolutions simultaneously.
The classification method is based on fuzzy logic with the membership functions based on shape, texture, hierarchy and relation to neighbor objects. Transmission
lines usually cross a variety of environments. The knowledge base constructed for
urban areas is shown is this paper.