Study of Methodology of Optimizing Image Segmentation and Processing Using Genetic Algorithm

Abstract

In the literature, several works focal point on the description of contrast metrics and standards that allow to count the overall an image processing algorithm performance. These assessment standards can be investigated to outline new photo processing algorithms by way of optimizing them. In this work, we suggest a universal scheme to phase photographs with the aid of a genetic algorithm. The developed approach makes use of an estimation standard which counts the first-class of a photo segmentation outcome. The suggested type segmentation approach can combine a local ground reality when it is accessible in order to set the preferred degree of the last result accuracy. A genetic algorithm(G-A) is then investigated in sort to decide the first-rate mixture of statistics excerpted with the aid of the chosen standard. Then, we exhibit that this method can either be utilized for gray-levels or multicomponent snap shots in a supervised context or in an unsupervised one. Last, we exhibit the efficiency of the suggested approach via several experimental outcomes on a number of (levels of gray) and multi-components images.