Genetic algorithm optimization based on support vector machine image interpolation

The fitting accuracy and generalization ability of support vector regression depends on the selection of its parameters. There have a strong dependence of the image correlation with image interpolation algorithm. Genetic algorithm has fast global search capability. So the genetic algorithm is applied to support vector machines parameters optimization, and the selected optimal parameters combination the image correlation applied to the image interpolation. Simulation results show that the proposed scheme produces visually pleasing images and obtains higher PSNR and SNR and smaller NMSE and MSE than other well-known image interpolation algorithms.