摘要
针对传统超分辨率算法存在的局部失真和块效应问题提出了一种改进的算法。算法以马尔科夫模型为基础,在满足图像重构的约束条件下,利用非线性局部搜索技术找到训练集中的最优分块,通过水平相容性检查来实现匹配块间的兼容性。基于Sigmoid函数进行加权处理,提高了匹配的精度。实验结果表明,该算法在获取HR图像的过程中能有效地抑制块效应和局部失真现象的出现,改善了超分辨率图像的质量,与传统算法相比,该算法的鲁棒性较强。
To mitigate the blocking ettect and local distortion m tradmtional super-resolution algorithms,an improved algorithm is proposed based on Markov model. This algorithm can meet the constraints of image reconstruction,obtain the optimal block from the train-ing set using nonlinear local searching technique,and achieve compatibility between matching blocks through horizontal compatibility chec- king. It also provides weighted processing based on Sigmoid function to improve matching accuracy. Experimental results demonstrate that the proposed algorithm can effectively prevent tbe blocking effect and local distortion in the process to obtain HR image,and improve the quality of the super-resolution image. Compared with the traditional algorithms, this algorithm also provides strong robustness.
出处
《无线电工程》
2012年第9期28-31,共4页
Radio Engineering
