Minimum Spanning Tree-based Structural Similarity Clustering for Image Mining with Local Region Outliers

Abstract

mining is more than just an extension of data mining to image domain. Image mining is a technique commonly used to extract knowledge directly from image. Image segmentation is the first step in image mining. We treat image segmentation as graph partitioning problem. In this paper we propose a novel algorithm, Minimum Spanning Tree based Structural Similarity Clustering for Image Mining with Local Region Outliers (MSTSSCIMLRO) to segment the given image and to detect anomalous pattern (outliers). In MSTSSCIMLRO algorithm we use weighted Euclidean distance for edges, which is key element in building the graph from image. MST-based image segmentation is fast and efficient method of generating a set of segments from an image. The algorithm uses a new cluster validation criterion based on the geometric property of data partition of the data set in order to find the proper number of segments. The algorithm works in two phases. The first phase of the algorithm creates optimal number of clusters/segments, where as the second phase of the algorithm further segments the optimal number of clusters/segments and detect local region outliers