Genetic Algorithm is a macro-heuristic algorithm based on the theory of natural selection. It uses genetic operators such as selection, crossover, and mutation to generate or search for higher-quality solutions in the population, so as to achieve the optimal solution to the problem.

 

Genetic algorithm can realize thermal infrared image enhancement well, and its essence is to find a gray scale transformation relationship. The genetic algorithm first encodes the gray-level correspondence between the output image and the input image, and randomly initializes the gray-level correspondence (that is, the first generation of individuals), then constructs a fitness function according to a certain image quality evaluation standard, and then repeatedly uses the genetic operation ( Including mutation, crossover, selection) evolution to generate new offspring individuals until the optimization criterion is met, so that the optimal or nearly optimal gray-scale transformation relationship is found, and this relationship is finally applied to achieve image enhancement, and the visible effect is more reasonable. think.

 

Constructing a reasonable fitness function is the key to obtain high-quality solutions by genetic algorithm. In addition, the genetic algorithm should pay attention to solving the problem of excessive calculation, because it may set a larger population size for the problem and go through a longer evolutionary process to search for the solution.