Comparative Study of Combination of Swarm Intelligence and Fuzzy C Means Clustering for Medical Image Segmentation
The image segmentation issues have been exploited by researchers over the years for diverse application. A hybrid algorithm for image segmentation is proposed in this paper which is the integration of fuzzy c means (FCM) clustering and swarm intelligence.
- PDF / 374,924 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 55 Downloads / 217 Views
Abstract The image segmentation issues have been exploited by researchers over the years for diverse application. A hybrid algorithm for image segmentation is proposed in this paper which is the integration of fuzzy c means (FCM) clustering and swarm intelligence. The algorithm is applied to segmentation problems of two medical image modalities, i.e., magnetic resonance imaging (MRI) image, and computed tomography (CT) image. A detailed comparison of the different swarm intelligence based algorithms is presented. The optimization technique is used to generate optimized cluster centers in the image segmentation process. The effectiveness of the algorithms is validated by cluster validity indices. Keywords Image segmentation
FCM Swarm intelligence MRI CT
1 Introduction Image segmentation is an active area of research related to digital image processing. It is an important stage of image processing for various applications which includes robotics, medical image segmentation, remote sensing, computer vision problems, etc. Recently, image processing has become an important element in medical diagnosis and development of an automated diagnosis system which is generally referred to as computer-assisted diagnosis (CAD). Image segmentation [1] is a method of identification and categorization of similar areas in an image or it can be considered as the separation of region of interest (ROI) in an image. Segmentation T. Ibungomacha Singh R. Laishram (&) Manipur Institute of Technology, Takyelpat, Manipur, India e-mail: [email protected] T. Ibungomacha Singh e-mail: [email protected] S. Roy Assam University, Silchar, Assam, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 A. K. Luhach et al. (eds.), Smart Computational Strategies: Theoretical and Practical Aspects, https://doi.org/10.1007/978-981-13-6295-8_7
69
70
T. Ibungomacha Singh et al.
helps in the detailed understanding of the image and it is an inseparable stage of image processing application which cannot be neglected. Hence more dedicated research is required in image segmentation. In this proposed work, we modeled image segmentation as a clustering problem and one of the most popular clustering algorithms, i.e., fuzzy c means (FCM) [2, 3] clustering algorithm is employed for image segmentation. Further swarm intelligence based optimization algorithms is integrated with the FCM algorithm to improve the segmentation. The combined algorithms are tested on two different medical images which are a brain MRI image and CT liver image. The algorithms are validated using clustering validity indices for its effectiveness. The population-based optimization techniques have been utilized in many engineering and nonengineering problems and prove to be quite successful. This idea generated lots of interest and guided to the development of many nature-inspired algorithms. In this paper, we have considered few of the popular algorithms such as genetic algorithms (GA), particle swarm optimization (PSO), artificial bee colony
Data Loading...
