Local binary pattern texture feature for satellite imagery classification

Abstract

The Texture Feature Extraction (TFE) plays an important role in satellite image processing application. This paper proposes a novel method for Satellite Imagery Classification. Our proposed method is a combination of Local Binary Pattern (LBP) and Fuzzy c-means classification algorithm. Local Binary Pattern is calculated by thresholding a 3 × 3 neighborhood of each pixel by the center pixel value. During the Feature Extraction Phase, Local Binary Pattern extracts the important characteristics from the satellite images. Fuzzy c-means algorithm classifying the images into different classes. This is a very challenging task in texture feature extraction being used in satellite images.