Enhancement of low level image processing methods using modified nature inspired metaheuristics

Abstract

Image enhancement and segmentation are significant and crucial low level image processing newlineprocedures which facilitate the high-level image processing techniques. However, proper enhancement newlineand segmentation of different kinds of images is a challenging task as both depend significantly on the newlineapplication requirements or objectives. Recently, to solve these issues, both image enhancement and newlinesegmentation are being formulated as optimization problems which are solved by employing nature- newlineinspired optimization algorithms with the assistance of different objective functions. These objective newlinefunctions are designed to assist the requirements of a given application. newlineIn this thesis, image enhancement and segmentation techniques have been proposed newlineconsidering gray and colour images using different nature-inspired optimization algorithms. For better newlinebrightness preserved image enhancement, modification of classical Histogram Equalization (HE) has newlinebeen performed to develop some HE variants namely 1. Bi-Thresholded Histogram based Variants, 2. newlineRecursive and Multi-Separation Histogram based Variants, 3. Weighted and Thresholded Histogram newlinebased Variants, 4. Range Optimized and Bi-Thresholded Histogram based Variants. newline