Texture Image Segmentation Using Affinity Propagation and Spectral Clustering

Abstract

Clustering is a popular and e®ective method for image segmentation. However, existing cluster methods often su®er the following problems: (1) Need a huge space and a lot of computation when the input data are large. (2) Need to assign some parameters (e.g. number of clusters) in advance which will a®ect the clustering results greatly. To save the space and computation, reduce the sensitivity of the parameters, and improve the e®ectiveness and e±ciency of the clustering algorithms, we construct a new clustering algorithm for image segmentation. The new algorithm consists of two phases: coarsening clustering and exact clustering. First, we use A±nity Propagation (AP) algorithm for coarsening. Speci¯cally, in order to save the space and computational cost, we only compute the similarity between each point and its t nearest neighbors, and get a condensed similarity matrix (with only t columns, where t << N and N is the number of data points). Second, to further improve the e±ciency and e®ectiveness of the proposed algorithm, the Self-tuning Spectral Clustering (SSC) is used to the resulted points (the representative points gotten in the¯rst phase) to do the exact clustering. As a result, the proposed algorithm can quickly and precisely realize the clustering for texture image segmentation. The experimental results show that the proposed algorithm is more e±cient than the compared algorithms FCM, K-means and SOM.

One-pass MapReduce-based clustering method for mixed large scale data

“…Examples of clustering methods categories are hierarchical methods, density-based methods, grid-based methods, model-based methods and partitional methods (Jain et al 1999). These clustering methods were well used in several applications such as intrusion detection (Tsai et al 2009;Wang et al 2010), customer segmentation (Liu and Ong 2008), document clustering (Ben N'Cir and Essoussi 2015; Hussain et al 2014), image organization (Ayech and Ziou 2015;Du et al 2015). In fact, conventional clustering methods are not suitable when dealing with large scale data.…”