08-01-2014, 04:23 PM
Object Extraction From High Resolution Satellite Images Using Texture Information
Abstract:
Texture in an image refers frequency of tonal variation. Now a day’s texture plays a vital role in various fields including pattern recognition, image segmentation image classification and machine vision. The purpose of the present text is to present the theory and techniques behind the first order texture method, and the state-of-the-art of the field, as applied to two dimensional images. In this paper, image objects are identified using four first order texture descriptors. Image objects are better identified using Mean and variance texture descriptor. Skew and Kurtosis fail to produce good results by adopting smaller window size, these two descriptor perform well with larger window size. The extracted texture descriptors can further be utilized in image classification process.