Radar Image Texture Classification based on Gabor Filter Bank

Mbainaibeye Jérôme; Olfa Marrakchi Charfi

The aim of this paper is to design and develop a filter bank for the detection and classification of radar image texture with 4.6m resolution obtained by airborne synthetic Aperture Radar. The textures of this kind of images are more correlated and contain forms with random disposition. The design and the developing of the filter bank is based on Gabor filter. We have elaborated a set of filters applied to each set of feature texture allowing its identification and enhancement in comparison with other textures. The filter bank which we have elaborated is represented by a combination of different texture filters. After processing, the selected filter bank is the filter bank which allows the identification of all the textures of an image with a significant identification rate. This developed filter is applied to radar image and the obtained results are compared with those obtained by using filter banks issue from the generalized Gaussian models (GGM). We have shown that Gabor filter developed in this work gives the classification rate greater than the results obtained by Generalized Gaussian model. The main contribution of this work is the generation of the filter banks able to give an optimal filter bank for a given texture and in particular for radar image textures.

Select Volume / Issues:

• Volume 5 Issue 1, Jan. 2014

Year:

2014

Type of Publication:

Article

Keywords:

Radar Image; Filter Bank; Gabor Filtering; Texture; Identification Rate; Classification Process

Journal:

IJECCE

Volume:

5

Number:

1

Pages:

184-196

Month:

January