超分辨率重建方法在图像处理中的应用

Applied research on supper resolution reconstruction method for image enhancement

为提升现有舰载光电探测系统图像的分辨率,通过图像降质及超分辨重建模型对复原图像中伪信息的产生进行了研究,在Lee方法的基础上,提出了采用高斯马尔科夫先验模型对图像重建过程进行约束的方法,实现了正则化参数的自适应计算。实验结果表明:该方法在视觉效果与客观指标上要优于传统方法与Lee方法,具有良好的应用前景。

To enhance the image resolution of the existing shipboard photoelectric detection system,this paper analyzes the pseudo image in image reconstruction by means of image degradation and super-resolution reconstruction model.To deal with the inadequacies of traditional methods,a Leebased method is proposed to use Gaussian Markov random field as a prior model to constrain the image reconstruction process for the adaptive computation of regular parameters.Experiment results show that the proposed method is superior to the traditional method and Lee′s method in visual effects and objective indicators and that this technique has a broad applying prospect.