基于改进样本块的数字图像修复算法研究 |
| |
引用本文: | 郭勇,王梅.基于改进样本块的数字图像修复算法研究[J].人天科学研究,2013(10):156-158. |
| |
作者姓名: | 郭勇 王梅 |
| |
作者单位: | [1]西安工程大学计算机科学学院,陕西西安710048 [2]西安交通大学城市学院科学与信息管理系,陕西西安710018 |
| |
摘 要: | 针对已有的基于样本块的纹理合成修复算法存在修复误差累积高的问题,重点研究了一种改进的基于样本块的数字图像修复算法。通过块匹配法与边缘驱动填充顺序的结合、全局搜索与局部搜索的结合,充分发挥修补过程中填充优先权的作用,有效地修补了图像受损区域的纹理和结构信息。仿真实验结果测试表明,与已有的传统算法相比,受损图像的修补更加完善,修复的误差累积得到了较好的改善。
|
关 键 词: | 图像修复 纹理合成 Criminisi算法 样本块 优先级 |
Research on Exemplar-Based Digital Improving Image Inpainting Algorithms |
| |
Abstract: | Due to the high cumulative error of texture synthesis repair algorithm based on the sample block at present, fo- cus on an improved digital image inpainting algorithm. By the combination of block matching method and edge driving fill- ing order,global searching and local area searching,giving fully filling priority in the repair process, repairing the texture and structure of the damaged area of the image efficiently. The results of simulation tests show that, compared with the existing traditional algorithm, the inpainting of the damaged image, accumulation of the repairing errors has been good im- provement. |
| |
Keywords: | Image Inpainting Texture Synthesize Criminisi Algorithm Exemplar Patch Priority |
本文献已被 维普 等数据库收录! |