Parallel Genetic Algorithm based Thresholding Schemes for Image Segmentation

Abstract

In this thesis, the problem of image segmentation has been addressed using the notion of thresholding.Since the focus of this work is primarily on object/objects background classification and fault detection in a given scene, the segmentation problem is viewed as a classification problem. In this regard, the notion of thresholding has been used to classify the range of gray values and hence classifies the image. The gray level distributions of the original image or the proposed feature image have been used to obtain the optimal threshold. Initially, PGA based class models have been developed to classify different classes of a nonlinear multimodal function. This problem is formulated where the nonlinear multimodal function is viewed as consisting of multiple class distributions.Each class could be represented by the niche or peaks of that class.Hence, the problem has been formulated to detect the peaks of the functions. PGA based clustering algorithm has been proposed to maintain stable sub-populations in the niches and hence the peaks could be detected. A new interconnection model has been proposed for PGA to accelerate the rate of convergence to the optimal solution. Convergence analysis of the proposed PGA based algorithm has been carried out and is shown to converge to the solution. The proposed PGA based clustering algorithm could successfully be tested for different classes and is found to converge much faster than that of GA based clustering algorithm.