AN IMAGE RETRIEVAL HIGH LEVEL SEMANTIC FEATURES USING NOVEL FUZZY ASSOCIATION

Abstract

Data mining was into existence since a long period of time but image mining took over since the recent years as it was found to be simpler even for non-technical users to retrieve their requirements in the form of images. High level image semantic representation techniques are based on the idea of developing a model of each object to be recognized and identifying image regions which might contain examples of the image objects. An image retrieval using high level semantic features is extraction of low level color, shape and texture characteristics and their conversion into high level semantic features using fuzzy production rules, derived with the help of an image mining technique. Transforming the low level texture characteristics into high level semantic features such as texture of Sky, Sea, Sunset, Beach and Building etc. is made by calculation the low level texture characteristic of a typical set of corresponding textures and finding the “cluster center” values which is used in the fuzzy production rules Current State of Image Mining Research, different Issues in Image Mining, and overview of Applications of Image Mining. The K-nearest neighbor algorithm is used to classify the image collection. The training dataset is selected so that it represents the various images of each class. By comparing the classification results of the Novel Fuzzy Association and normal histogram representations, the Novel Fuzzy Association to represent the image data, improves the classification results as com- pared with the normal histogram representation of image data and in this dissertation found that Novel Fuzzy Association is more accurate than the normal histogram representation. It is also obvious that using approximations of the image data not only improves the quality of classification and clustering, but also significantly improves the efficiency of classification and clustering. MATLAB provides an intuitive language and flexible environment for technical computations which integrates mathematical computing and visualization tools for data analysis and development of algorithms and applications. MATLAB was first adopted by researches and practitioners in control engineering. It is now also used in education, in particular the teaching of learner algebra and numerical analysis and is popular amongst scientist involved in image processing. MATLAB has structured Syntax, Variable and Vectors / matrices. MATLAB supports structure data types. Since all variables in MATLAB are arrays, a more adequate name is “structure only”. Statistics further analysis, filtering, optimization, and numerical integrations 2-D and 3-D graphics functions for visualizing data. It is possible to develop a prototype of an application for a relatively