盲环境下数字图像的自动修复算法研究*

当前所有的数字图像修复算法都是一种半自动的方法，因为在对图像修复前，都是由用户给出想要修复的区域，而算法无法自动获取照片的破损所在。提出了盲环境下图像自动修复的概念，以基于数字图像的样本修复为基础，改进了基于纹理合成的图像修复算法。首次运用数字图像可信性评估体系于图像修复，并通过可信性综合度量模型判断确定待修复区域。实验证明，该算法最终达到了盲环境下的图像自动修复效果。

Research on arithmetic of automatic inpainting for digital image in blind environment

At present all the digital image restoration algorithm is a semi-automatic way, because the image before restoration, is given by the user wants to repair the region, and the algorithm cannot automatically get a photo where the damage. This paper presented a concept of digital image impainting, and based on exemplar-based image inpainting, improved image inpain-ting algorithm based on texture synthesis. This article was first assessment of the credibility of the use of digital imaging system for image inpainting, and trustworthiness synthesis measuring model through to judge the credibility of identified areas to be repaired. Experiments show that the algorithm eventually reaches the effect of the automatic blind image inpainting.