An Optimized Feature Selection for Image Classification Based on SVM-ACO

Multi- class classification plays an import role in image classification. Multi-class classification used different classifier for the classification of data, such as binary classifier and support vector machine. In this dissertation we proposed a feature sampling technique of image classification. Our sampling technique optimized the feature selection process and reduced the unclassified region in multi-class classification. For the process of optimization we used ant colony optimization algorithm for the proper selection of feature sub set selection. Ant colony optimization is very famous meta-heuristic function inspired by biological spices. For the classification of image data used support vector machine. Support Vector Machines are designed for binary classification. When dealing with several classes, as in object recognition and image classification, one needs an appropriate multi class method. Different possibilities include: Modify the design of the SVM, as in order to incorporate the multi-class learning directly in the quadratic solving algorithm. Combine several binary classifiers: “One-against- One” (OAO) applies pair wise comparisons between classes, while “One-against-All” (OAA) compares a given class with all the others put together. OAO and OAA classification based on SVM technique is efficient process, but this SVM based feature selection generate result on the unclassified of data. When the scale of data set increases the complexity of preprocessing is also increases, it is difficult to reduce noise and outlier of data set. Ant Colony Optimization (ACO) meta-heuristic is an effective tool in finding quality data and that’s the main reason to use it as a feature selection for SVM.