Title :
Superpixel-based classification of SAR images
Author :
Arisoy, Sertac ; Kayabol, Koray
Author_Institution :
Elektron. Muhendisligi Bolumu, Gebze Teknik Univ., Gebze, Turkey
Abstract :
We propose a superpixel-based classification method for finding area such as urban, land, lake etc. in SAR image. For classification, we use superpixel histograms as features. We compare the results of state-of-the-art superpixel methods with the one of recently proposed SAR image specific superpixel method. The test results obtained from a real TerraSAR-X image show that the SAR image specific method outperforms the state-of-the-art ones.
Keywords :
feature extraction; image classification; image resolution; radar imaging; synthetic aperture radar; SAR images; TerraSAR-X image; superpixel histograms; superpixel methods; superpixel-based classification method; Computer vision; Image segmentation; Nakagami distribution; Radar imaging; Remote sensing; Synthetic aperture radar; SAR image; Superpixel classification; finite mixture models;
Conference_Titel :
Signal Processing and Communications Applications Conference (SIU), 2015 23th
Conference_Location :
Malatya
DOI :
10.1109/SIU.2015.7129945