Title :
A multi-phase clustering method for PolSAR image segmentation
Author :
Akbarizadeh, Gholamreza ; Rahmani, Masoumeh
Author_Institution :
Dept. of Electr. Eng., Shahid Chamran Univ., Ahvaz, Iran
Abstract :
In this paper, a new multi-phase clustering method is proposed for PolSAR image segmentation. In this method, the PolSAR image is first segmented with mean-shift method in color space. Then, a texture feature vector for each pixel of the image is formed. Then, a new spectral clustering method is used to combine previous steps. The results show that the proposed method is effective in PolSAR image segmentation.
Keywords :
feature extraction; image colour analysis; image segmentation; image texture; pattern clustering; radar imaging; radar polarimetry; spectral analysis; synthetic aperture radar; PolSAR image segmentation; color space; mean-shift method; multiphase spectral clustering method; texture feature vector; Conferences; Decision support systems; Electrical engineering; Erbium; Oceans; Remote sensing; PolSAR image; spectral clustering; texture feature;
Conference_Titel :
Electrical Engineering (ICEE), 2015 23rd Iranian Conference on
Conference_Location :
Tehran
Print_ISBN :
978-1-4799-1971-0
DOI :
10.1109/IranianCEE.2015.7146232
