针对BSCB模型速度慢、整体变分模型易产生阶梯效应的缺点,提出了一种基于扩散率函数的自适应插值算法。该算法利用待修复像素点周围的已知像素点的梯度权函数和距离权函数,对未知像素点进行赋值。实验结果显示: 将该算法应用到灰度图像和彩色图像的修复中,与传统算法相比所需运算时间短,视觉效果较好,定量评价指标(峰值信噪比)也证明了该算法的有效性。
Abstract
Since the BSCB model has an inherent slow rate, and the total variation model is subject to staircase effect, to overcome these disadvantages, an adaptive interpolation algorithm is proposed based on the diffusivity function. The proposed algorithm repairs the unknown pixels by using the gradient weight function and distance weight function of the neighboring pixels whose values are known. Experimental results show that compared to the traditional methods, the proposed algorithm can obtain relatively higher image quality in a shorter time for both gray and color image restoration, additionally, the quantitative evaluation indicator has also verified the effectiveness of the proposed method.
关键词
计量学 /
图像修复 /
偏微分方程 /
扩散率函数; /
自适应插值
Key words
Metrology /
Image restoration /
Partial differential equation /
Diffusivity function /
Adaptive interpolation
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］ Li X D, Du Y, Ling F. Using a subpixel mapping model to improve the accuracy of landscape pattern indices［J］. Ecological Indicators,2011,11(5): 1160-1170.
［2］ 张红英. 数字图像修复技术的研究与应用［D］. 成都: 电子科技大学, 2006.
［3］ Criminisi A, Perez P, Toyama K. Objectremoval by exemplarbased inpainting［C］//Computer Vision and Pattern Recognition. USA: IEEE Computer Society Conference, 2003:721-728.
［4］ Shen Z Q, Qi J G, Wang K. Modification of pixel-swapping algorithm with initialization from a sub-pixel spatial attraction model［J］.Photogrammetric Engineering and Remote Sensing, 2009, 75(5) : 557-567.
［5］ 许威威,潘志庚,张明敏. 一种基于整体变分的图像修补算法［J］. 中国图象图形学报, 2002, 7(4): 351-355.
［6］ 江雯,陈更生,杨帆. 基于Sobel 算子的自适应图像缩放算法［J］. 计算机工程, 2010, 36(7): 214-216.
［7］ Hwang J W, Lee H S. Adaptive Image Interpolation Based on Local Gradient Features［J］. IEEE Signal Processing Letters, 2004, 11(3): 359-362.
［8］ 许微. 基于偏微分方程的图像修复及放大算法研究［D］. 天津大学, 2007.
［9］ Perona P, Malik J. Scale-space and edge detection using anisotropic diffusion［J］. IEEE Trans on Pattern Analysis and Machine Intelligence, 1990, 12(7): 629-639.
［10］ Lin S Y, Shi J Y. Fast natural image matting in perceptual color space［J］. Computers and Graphics, 2005,29(3): 403-414.
基金
国家自然科学基金青年基金(61102096)
PDF(736 KB)
