基于量子行为粒子群优化算法的图像插值方法

Image interpolation algorithm based on quantum-behaved particle swarm optimization

传统图像插值方法简单,容易实现,但经过插值后的图像会增加一定的虚假内容,导致图像模糊。为提高插值图像的质量和图像的分辨率,提出一种基于量子行为粒子群优化(QPSO)算法的图像插值方法。该方法利用QPSO算法在以传统插值图像为基础形成的解空间中,寻找符合目标函数的最优高分辨率图像。实验证明,该方法实用、可行,且能得到质量较好的插值图像。

The conventional interpolation algorithms of image are easy to be realized, but they resuh in high frequency artifacts in the interpolated image. In order to improve the quality of the interpolated image and enhance the resolution of it, an image interpolation algorithm based on Quantum-behaved Particle Swarm Optimization （QPSO） algorithm was proposed in this paper. This method uses QPSO algorithm to seek the best high resolution image through the objective function in the traditional interpolation image solution space. The experiments demonstrate that the proposed algorithm not only is practical and applicable, but also improves the quality of the interpolated images.