Title :
Extended Kalman filter based semi-automatic robust road detection
Author :
Karaomeroglu, Betul ; Baseski, Emre
Abstract :
The research on automatic road detection from satellite images started in the mid-70s [1]. Nowadays, there are many algorithms about non-supervised and semi-supervised road extraction. While some methods only use weak geometric information in combination with radiometric and spectral characteristics of the roads, more advanced local and global geometric properties such as smooth continuation, maximum curvature or network formation started to appear in recent algorithms.
Keywords :
Kalman filters; artificial satellites; feature extraction; geophysical image processing; nonlinear filters; object detection; radiometry; remote sensing; roads; extended Kalman filter; global geometric property; nonsupervised road extraction; radiometric characteristics; satellite images; semi-automatic robust road detection; semi-supervised road extraction; spectral characteristics; Equations; Feature extraction; Kalman filters; Mathematical model; Radiometry; Roads; Satellites;
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2014 IEEE International
Conference_Location :
Quebec City, QC
DOI :
10.1109/IGARSS.2014.6947595
