Unsupervised Texture-Based SAR Image Segmentation Using Spectral Regression and Gabor Filter Bank
- PDF / 1,362,006 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 113 Downloads / 181 Views
RESEARCH ARTICLE
Unsupervised Texture-Based SAR Image Segmentation Using Spectral Regression and Gabor Filter Bank Zeinab Tirandaz 1 & Gholamreza Akbarizadeh 1
Received: 9 March 2015 / Accepted: 10 August 2015 # Indian Society of Remote Sensing 2015
Abstract Segmentation of synthetic aperture radar (SAR) image is a difficult task in remote sensing applications due to the influence of the speckle noise. Most existing clustering algorithms suffer from long run times. A novel unsupervised segmentation algorithm has been proposed in this paper, based on Gabor filter bank and unsupervised spectral regression (USR), for SAR image segmentation. In the proposed algorithm, we use a Gabor filter bank to decompose the image to several subimages. Features are extracted from these sub-images and further, learned, using USR. Finally k-means clustering is employed and the image is segmented. The segmentation results were tested on simulated and real SAR images, texture images, and natural scenes. The results of segmentation on texture images show that proposed algorithm has the ability to effectively manage large-size segmentation cases, since the eigen-decomposition of the dense matrices is not required in USR. Meanwhile, the proposed algorithm was more accurate than all of the other compared methods. The running time in MATLAB was compared against parallel sparse spectral clustering (PSSC) and although our proposed algorithm is serial, it had significantly shorter run time compared to PSCC. It is also demonstrated that the clustering of features improves significantly after learning. Keywords Synthetic aperture radar (SAR) . Unsupervised spectral regression (USR) . Gabor filter bank . Texture segmentation . Clustering * Zeinab Tirandaz [email protected] Gholamreza Akbarizadeh [email protected] 1
Department of Electrical Engineering, Faculty of Engineering, Shahid Chamran University, Ahvaz, Iran
Introduction Synthetic aperture radar (SAR) is widely used in all areas of military, land-use classification, disaster aftermath analysis to estimate losses, oil spill detection, road extraction and so on, because of its ability to operate in day and night and all weather conditions. Unlike segmentation in optical images, such a task on SAR image is very difficult because of the existence of speckle noise (Akbarizadeh 2012; Peng et al. 2012). Texture-based segmentation is very important in understanding and processing SAR images. Feature extraction is a difficult task in SAR images because of the presence of speckle noise. Feature extraction should be performed carefully in these images and the best physical features of the images should be extracted (Jain and Farrokhnia 1991). We used Gabor filter bank to decompose the image. Local optimization in spatial/frequency domain, simplicity and the ability to simulate 2D behavior of simple cells in the visual cortex motivates us to use Gabor filter bank in this step. The simulation is accompanied with separating certain frequencies and orientations. There are two main phases wh
Data Loading...
