Sythetic Aperture Radar Image Change Detection Using Swt and Flicm Algorithm

Synthetic Aperture Radar Image Change Detection using SWT and FLICM Algorithm

Dimina .K.C

Department of Electronics and Communication

Meenakshi Sundararajan Engineering College,

Chennai, India

Maithreyi Rajasekar

          Department of Electronics and Communication

Meenakshi Sundararajan Engineering College,

Chennai, India

Abstract-- Image change detection is a process in which two or more images of same scene that are captured at two different time instants are taken and changes are detected. Out of many categories of images, detection of changes in Synthetic Aperture Radar (SAR) images is a critical task because of the speckle noise present in these images. In this paper changes are detected in SAR images based on an image fusion strategy and fuzzy clustering algorithms. Image fusion is used to generate a difference image from the information obtained from mean ration and log ratio operator and Stationary Wavelet Transform (SWT) is used to decompose the sub bands and the fuzzy local information C means clustering (FLICM) algorithm is applied on the fused image and the changed and unchanged areas are differentiated by forming clusters of similar and dissimilar elements. It incorporates the information about spatial context in a novel fuzzy way for the purpose of enhancing the changed information and reducing the effect of speckle noise.

Keywords— Clustering, fuzzy C-means (FCM) algorithm,

Image change detection, image fusion, Stationary Wavelet Transform(SWT),synthetic aperture radar(SAR).

1. INTRODUCTION

      Image change detection is a process that analyzes a pair of images which are acquired on the same geographical area at two different time instants. It has found important applications in remote sensing [1], flood monitoring, geographical map updating [2], medical diagnosis and video surveillance [3]. Remote sensing systems offers many advantages out of which one is the capability of repetitive coverage of a particular area at different times to study change detection process. Different types of sensors can be used to accomplish this task. Synthetic aperture radar (SAR) imagery [4] has found a major field of interest when compared to optical ones, due to its advantages such as operating in various weather conditions and are not affected by different sunlight conditions and cloud cover etc., but these SAR images contain some unwanted artifacts in the  form of granules which is called as speckle noise. Due to the presence of this speckle noise, some of the crucial and important information of the images may be lost. In spite of its drawback change detection is clearly done due to its advantages. Many algorithms have been already proposed to detect the changes in SAR images.