A Survey on Image Contrast Enhancement Using Genetic Algorithm

This paper reviews an introduction to various approaches to image contrast enhancement in the spatial domain using genetic algorithm and its extension based on population based incremental learning (PBIL). GA performs efficient search in global spaces to get an optimal solution. The algorithm does not require any prior knowledge about image in Order to select the appropriate enhancement function. GA is more effective in the contrast enhancement and produce image with natural contrast. Histogram equalization and similar methods for image contrast enhancement produce unnatural brightness. This paper introduces various approaches based on genetic algorithm to get image with good and natural contrast. genetic algorithm is a type of search algorithm that takes input and computes an output where multiple solutions might be taken. It is a mechanism based on natural selection and natural genetic. It works well in global search space. Genetic algorithm uses the principle of selection and evolution to produce solutions at each generation. In simple genetic algorithm the size of whole population is same (2). It uses strings to represent a chromosome. Genetic algorithm works as - The initial population of solutions is created by a group of individuals randomly. These candidate solutions are called chromosomes. The individuals in the population are evaluated using a fitness function to measure the work of chromosome towards solving the problem. Two individuals are selected based on their fitness, the higher fitness, and the higher chance of being selected. Only fittest individuals are allowed to survive for next generation (1). These individuals perform crossover to reproduce one or more offspring using crossover function. Some individuals are mutated randomly. Basically genetic algorithm used in any study are characterised by parameters- population size, selection, crossover, type, crossover rate and mutation rate. The population size shows the number of chromosomes that are present in every generation. Selection is used to select the individuals for next generation. The crossover type is used to get a way to recombine information. Crossover is used to recombine two strings to get better string. The two strings that are involved in the crossover operation are known as parent string and resulting strings are known as child strings. Crossover operation is done by randomly selecting two strings for crossover operation. Many crossover operators exist in genetic algorithm. Single point crossover, a point of exchange is selected randomly in the two individual's genomes and swaps the content of chromosomes to produce an offspring's. Two point crossover, two crossover points are selected randomly and swaps the contents of chromosomes to produce an offspring's. Uniform crossover, the value at any given location in the offspring's is either the value of genome of one parent at that position or the value of genomes of the other parent at that position. Following exemplify the crossover process.