Title :
Power line detection using a circle based search with UAV images
Author :
Ceron, Alexander ; Mondragon B, Ivan F. ; Prieto, Flavio
Author_Institution :
Fac. of Eng., Univ. Mil. Nueva Granada, Bogota, Colombia
Abstract :
A new method for power line detection based on computer graphics algorithms is presented. The algorithm uses geometric relationships that are inherent to the circle symmetry. The method detects line segments that are linked in a posterior stage. For the detection, we use Canny and Steerable Filters. We developed two tests for validating the proposed approach. The first one uses synthetic images and the second one real power line images taken from UAVs. The results show that this method is not only efficient for line detection, but it also takes compared with state of the art algorithms a short computing time without the use of a GPU approach.
Keywords :
autonomous aerial vehicles; computer graphics; filtering theory; image segmentation; object detection; power cables; power engineering computing; Canny filters; UAV images; circle based search; circle symmetry; computer graphics algorithms; line segment detection; power line detection; power line images; steerable filters; Computer graphics; Detectors; Image edge detection; Image segmentation; Joining processes; Object segmentation; Transforms;
Conference_Titel :
Unmanned Aircraft Systems (ICUAS), 2014 International Conference on
Conference_Location :
Orlando, FL
DOI :
10.1109/ICUAS.2014.6842307