一种优化修复顺序的图像修复算法

在基于样本的纹理合成的图像修复算法中,缺损区域的填充顺序至关重要,极大的影响着图像的修复质量。在A.Criticism算法的基础上,本文通过引入递增权值算子λ,增强修复后期结构信息量D(p)在确定缺损区域边界像素的修复优先权时的作用,从而得到更好的,更合理的修复顺序。对大量的较大面积的缺损图片进行仿真后的结果表明,本文提出的改进算法能够取得比A.Criticism算法更好的修复效果。

One kind of image initiating algorithm that optimize initiating order

In texture synthesis image initiating algorithm based on the sample,the initiating order of the damaged area is very important,which influences image initiating quality enormously.Based on A.Criticism algorithm,this article enhances the function of structure information content D(p) in determining the initiating priority of the pixel which is on the boundary of damage region through introducing increasing weight operator,thus obtains the better,more reasonable initiating order.The result after carrying on the simulation of massive images with wide damaged area indicates that the improved algorithm proposed by this article can obtain a better initiating effect compared to A.Criticism algorithm.