摘要
图像修复是对图像的破损区域的复原,实现对图像的小区域的重建。对图像修复的主要方法有两类:一种是基于偏微分方程(PDE)的修复方法,另外一种为基于纹理合成的修复算法。大多数现有图像修复算法都非常复杂,修复速度慢。基于纹理合成算法主要是搜索的范围太大导致效率低,为了在修复效果和效率之间取得折中,文中提出了一种简单的基于块匹配修复算法,应用不同的修复模板,根据图像的空间连续性,缩小了匹配块的搜索范围,大大提高了修复效率。该算法实现简单,修复速度快,对空间较连续的图像修复效果较好。
Digital inpainting provides a means for reconstruction of small damaged portions of an image.Although the inpainting basics are straightforward,now the image inpainting have two main kinds：one based on partial differential equations（PDEs）and the second based on texture synthesis.Most inpainting techniques published in the literature are complex to understand and implement.The exhaustive searching strategies used in previous work are the main source of inefficiency,in order to make a tradeoff between time and quality,propose a simple image inpainting algorithm based on texture synthesis and spatial continuity,use different masks,searching local area.The algorithm is very simple to implement,fast and can produce comparable results.
出处
《计算机技术与发展》
2010年第11期80-83,共4页
Computer Technology and Development
基金
安徽省自然科学研究项目(KJ2010B296
KJ2009B121Z)
安徽科技学院引进(稳定)人才专项基金(ZRC2008180
ZRC2008179
ZRC2008181)
