Segmenting of Remote Sensing Images into Different land Cover Types using Otsu Thresholding

Abstract

Image segmentation is an important initial operation before building databases for GIS use. Otsu method is a fast and efficient thresholding approach because it is minimizing within class variance. In this study image segmentation has been done by using Otsu thresholding. First, an image contains two main cover types processed by Low Pass filter to enhance separation between classes.Then, use thresholding to segment image into distinct regions (classes), water and land cover. Finally, each distinct region will store in binary image to uses in later image processing. Keywords: Segmenting of remote sensing, thresholding approach, binary image.