论文部分内容阅读
针对局部自相似性重建方法的块效应问题,以及MRF网络模型方法外部训练库数据不相关性产生的图像重建误差问题,提出了一种结合局部自相似性和MRF网络模型的超分辨率重建方法。首先,利用图像局部自相似特性,引入自身冗余信息构建高分训练库,然后建立低分与高分训练库映射的MRF网络模型,通过置信传播算法求解MRF模型重建高分图像。以仿真和真实卫星图像进行超分实验,结果表明本文方法能够改善图像的细节,较好地去除了块效应,提高了地物边缘的清晰度。
Aiming at the block effect of local self-similarity reconstruction and the image reconstruction error caused by the irrelevance of MRF network model training data, a super-resolution reconstruction method combining local self-similarity and MRF network model is proposed . First of all, using the local self-similar feature of image, introducing the redundant information to construct the high score training library, then constructing the MRF network model mapping the low score and high score training library, and using MRF model to reconstruct the high score image. Experimental results show that the method proposed in this paper can improve the detail of the image, remove the block effect and improve the sharpness of the edge of the object.