Title :
Buried pipeline mapping from satellite images
Author :
Petrou, Maria ; Gracia, Isabel
Author_Institution :
Sch. of Electron., Comput. & Math., Surrey Univ., Guildford, UK
fDate :
8/1/2002 12:00:00 AM
Abstract :
A semi-automatic method for the detection and tracking of buried pipelines under a variety of surface coverages is presented and its performance in experiments reported. In high 1 m resolution aerial or satellite images, the buried structures manifest themselves as bands a few pixels wide having contrast and texture changes of the overground growth. Some statistical nonlinear filters are proposed for the detection of such wide linear features, and they are used for the detection of pipeline routes. The pipeline routes are then tracked by exploiting their lateral continuity
Keywords :
buried object detection; feature extraction; filtering theory; image resolution; nonlinear filters; statistical analysis; aerial images; buried pipeline detection; buried pipeline mapping; buried pipeline tracking; buried structures; contrast changes; feature detection; image resolution; pixels; satellite images; semi-automatic method; statistical nonlinear filters; surface coverages; texture changes;
Journal_Title :
Electronics & Communication Engineering Journal
DOI :
10.1049/ecej:20020402