纹理细节和边缘结构保持的图像插值算法

针对图像重建过程中产生的边缘结构被破坏和纹理细节丢失问题,提出一种纹理细节和边缘结构保持的图像插值算法。首先,采用自适应阈值的八方向边缘检测划分图像区域;其次,构造双变量有理函数模型,它可在有理模型和多项式模型之间转换;最后,提出基于边缘数据的局部不对称性和梯度特征调整待插值点空间距离的方法,调整边缘部分待插值点坐标并代入有理模型实现插值,而非边缘部分采用多项式模型插值。实验表明,该算法的峰值信噪比平均提升了0.48~2.17 dB,结构相似性平均提升了0.004~0.028,获得了较高的客观评价数据。该算法将原空间距离不变的插值修正为空间距离变化的插值,有效地保持了图像的边缘结构和纹理细节,使得重建结果具有较好的视觉效果。

Image interpolation algorithm based on texture details and edge structure maintained

In order to solve the problem of destroys the edge structure and losses the texture detail in the process of image reconstruction,this paper proposed an image interpolation algorithm for preserving the texture detail and the edge structure.Firstly,it divided image area by using eight direction edge detection based on adaptive threshold.Secondly,it constructed a bivariate rational function model,which could convert into rational model and polynomial model.Finally,based on the local asymmetry and gradient feature of the edge data,this paper proposed a method to adjust the distance of points to be interpolated.By this method,it adjusted the coordinates of the points to be interpolated and used rational model to realize interpolation,as for the non-edge part,it adopted polynomial model to interpolation.Experiments show that,the peak signal to noise ratio of the algorithm increases 0.48~2.17 dB on average,and the structural similarity index increases 0.004~0.028 on ave-rage.This algorithm obtains high objective evaluation data.The proposed algorithm modifies the original spatial distance inva-riant interpolation to the spatial distance varying interpolation,which effectively keeps the edge structure and texture details of the image,so that the reconstruction result has better visual effect.