摘要
提出了一种基于自适应块匹配的超分辨重建算法。首先根据相邻运动块之间的相关性来预测搜索块,然后利用图像的运动特性自适应地选择合适的搜索模式,最后采用最大后验概率估计算法对配准后的图像进行超分辨率重建。该算法可以有效减少搜索匹配点的个数,极大提高搜索速度,并且能够获得与全搜索算法同样高的配准精度,具有较大的峰值信噪比。实验结果表明,所设计算法能获得较好的视觉效果与重建效果,具有一定的实用价值。
This paper presents an adaptive block matching algorithm of super-resolution reconstruction. Firstly,the search block is predicted according to the correlation between adjacent blocks,and then the appropriate search mode is adaptively selected by using the motion characteristics of the image,finally,the image after rectification is super-resolution reconstructed by the maximum posterior probability estimation algorithm. The algorithm greatly improves the search rate,at the same time effectively reduces the search number of matching points,and can obtain the same registration accuracy compared with the full search algorithm,maintains a high peak signal to noise ratio（ PSNR）. Experimental results show that the algorithm has better visual effect and reconstruction effect,which confirms a practical value of the algorithm.
出处
《四川理工学院学报（自然科学版）》
CAS
2015年第3期33-36,共4页
Journal of Sichuan University of Science & Engineering(Natural Science Edition)
基金
四川省教育厅科研项目(13ZB0138)
人工智能四川省重点实验室开放基金项目(2013RYY02)
关键词
超分辨率重建
自适应块匹配
最大后验概率
峰值信噪比
super-resolution reconstruction
adaptive block matching
maximum posterior probability
PSNR
