方向性纹理图像的修复算法  

Algorithm of Completion for Strong Orientational Texture Images

作　　者：汪忠庆[1] 檀结庆[2] 何蕾[2] 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009 [2]合肥工业大学理学院,合肥230009

出　　处：《计算机工程与应用》2006年第34期29-32,共4页Computer Engineering and Applications

基　　金：国家自然科学基金资助项目(10171026;60473114);安徽省自然科学基金资助项目(03046102)。

摘　　要：自然图片和一些方向性纹理图像通常有污点或者希望去掉覆盖背景的一些目标,如果直接擦掉前景会留下一个“空洞”。针对这种强方向性纹理分布图像,提出一种修复算法去填充这些“空洞”。由于强方向性,算法通过在局部区域选择图像块以合成缺失的图像块。若缺失信息的图像块BT附近的为已知图像块BTN,沿着纹理/颜色分布的方向或者约束的方向搜索,可以找到与BTN最相似的块BSN,并用BSN附近的图像块BS去更新BT。在更新BT前,计算块与块之间边界的最小误差路径,以其为更新边界,以减小拼接的痕迹。最后文章给出了对不同的纹理图像修复的例子,以说明方法的有效性。Natural photographs and strong orientational texture images sometimes may contain stains or undesired objects covering significant portions of the images,where the removal of a foreground object creates a hole in the images.We present an inpainting algorithm for strong oricntational texture/color images existing in our environment to fill in the holes.Becauce of the strong orientation,the algorithm synthesizes the missing part by image patches selected from orientationally local areas.Along the orietation of texture /color distribution or the constrained orietation to search,we select a source block BTN near a missing block Br firstly,and find the most similar block BSN to BTS ,then update Br by the block BS near BSN.To reduce blockiness,the boundary between blocks is further computed as a minimum cost path through the error surface at the overlap before updating Br .A few examples for a variety of texture images are given to demonstrate the effectiveness of our method.

关 键 词：图像修复 纹理合成 目标去除 最小误差路径 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]