Analysis of Noisy Satellite Image Using Statistical Approach

Abstract

Technological advances and increasing availability of satellite sensors acquire more information about the earth and offer the potential for more accurate land cover classifications and pattern analysis. However, this type of image (satellite image) is rich and various in content, however it suffers from noise that affects the image in the acquisition. Therefore, there is a requirement of an effective and efficient method for features extraction from the noisy image. This paper presents an approach for satellite image segmentation that automatically segments image using a supervised learning algorithm into urban and nonurban area. We have applied a statistical feature including local feature computed by using the probability distribution of the phase congruency computed (El Fellah S, El haziti M, Rziza M, et Mastere M, A hybrid feature extraction for satellite image segmentation using statistical global and local feature, 2016, [1]). The results provided, demonstrate a good detection of urban area with high accuracy in absence of noise. However when noise is added to images, the classification results deteriorate. Hence, to improve these results we propose a novel features based on higher order spectra known by their robustness against noise.