Detection in SAR images based on histogram and improved elitist genetic fuzzy clustering

Abstract

Change detection in Synthetic Aperture Radar Images has been an important technique for Synthetic Aperture Radar Images. In this paper, a novel unsupervised change detection algorithm based on histogram and improved elitist genetic fuzzy clustering is proposed. First, a difference image is generated by multiplying transform fusion. Second, we use the characteristics of the histogram to deal with the difference image. Then, the new algorithm is proposed to partition these characteristics into changed and unchanged regions. The proposed algorithm has the following merits: 1. FCM is employed to initialize the population and to calculate the fitness function of the genetic algorithm. 2. The optimal solution is selected by an elitist selection strategy based on population concentration and the optimal solution will be the initial clustering center of FCM, which significantly increases the convergence speed. 3. The histogram is utilized to reduce the sample points of images. Compared with the state-of-the-art algorithms, the experimental results demonstrate the effectiveness in processing of change detection in SAR images.