基于改进TV模型的修复算法在岩画修复中的应用

Application of Image Repair Algorithm Based on Improved TV Model in Rock Painting Repair

从数学角度来看,图像修复就是将待修补区域周围的信息填充到待修补区域中。传统的基于TV模型的图像修复算法对小区域破损的图像有较好的修复效果,但是对参数较敏感,传统TV算法修复稍大破损范围的图像容易出现模糊现象和阶梯效应。为了克服上述缺陷,结合非线性扩散的思想提出了一种改进的基于TV模型的岩画修复算法,该算法能够有效解决传统TV修复算法小区域失效的问题。实验结果表明,该算法相较于传统的TV模型取得了较好的修复效果。

Mathematically,image restoration is to fill the area around the information to be filled into the area to be patched.The traditional image restoration algorithm based on TV model has better repair effect on the image of small damaged area.But it is more sensitive to the parameters,and is prone to produce fuzzy phenomenon and ladder effect in the range of slightly large damaged image.In order to overcome the above defects,an improved rock model of repair algorithm based on TV model is proposed combined with the idea of nonlinear diffusion in this paper,which can effectively solve the failure problem of traditional image restoration algorithm in small area.The experimental results show that the proposed algorithm works better than traditional TV model.