摘要
将压缩感知理论应用于图像超分辨率重建技术,并针对总变分正则化方法未能考虑图像局部纹理特性的缺点,提出一种基于局部方向特性的总变分正则化函数.通过在图像的有界变差函数空间中构造与图像纹理方向一致的椭圆形区域,再利用最小化算法进行迭代求解获得去噪图像,可以达到更好地重建图像边缘信息的效果.实验结果表明,该算法可以有效降低图像的均方误差,提高峰值信噪比,并在重建过程中保持了图像边缘和纹理细节,提高了图像的对比度和清晰度.
In this paper compressive sensing theory is applied to image super-resolution reconstruction technique,and in order to overcome the disadvantage that the total variation regularization method without considering image's local texture feature,an adaptive total variation function based on local orientation features is proposed.Construct an ellipse region coincides with the dominant direction of the local image in the space of the bounded variation function.Then to obtain the denoised image by using iteration to optimize the minimum total variation.Experimental results show that the proposed algorithm can effectively reduce the mean square error of the image,improve the peak signal to noise ratio,and maintain the image edge and texture details in the reconstruction process,so as to improve the contrast and clarity of the image.
出处
《内蒙古师范大学学报(自然科学汉文版)》
CAS
北大核心
2016年第3期397-402,共6页
Journal of Inner Mongolia Normal University(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
压缩感知
局部纹理
总变分正则化
图像去噪
super-resolution reconstruction
compressive sensing
local texture
total variation
image denosing