【作者】 沈妮； 张红英； 彭启琮；

【Author】 SHEN ni,ZHANG hong-ying,PENG qi-cong Lab 140, School of Communication and Information Engineering, UESTC (School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, 610054)

【机构】 电子科技大学通信与信息工程学院；

【摘要】 数字图像补全是图像处理的研究热点。传统的纹理合成算法采用固定大小的匹配块进行搜索,本文提到的算法是一种新型的基于纹理合成的图像补全算法,该算法自适应计算匹配块大小,提高了修复质量。文章通过将算法在DSP上做了实现,发现用DSP实现图像修复比传统的修复方法速度快、效果好、操作简单,而且可以通过调节相关参数,达到不同的处理效果。更多还原

【Abstract】 Image inpainting is an important research topic in the area of image processing. The traditional texture synthesis algorithms use fixed matching template as the searching sample. And the one mentioned by this paper is an advanced one. DSP is widely used in digital signal processing. This paper proposed a new inpainting method by combine the two. Some experiment results show that the image inpainting on DSP is feasible, quick and easy to operate, and different inpainting results can be attained by adjusting certain parameters.更多还原