• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International Journal of Innovative Computing Information and Control >A NEW INPAINTING METHOD FOR OBJECT REMOVAL BASED ON PATCH LOCAL FEATURE AND SPARSE REPRESENTATION
【24h】

A NEW INPAINTING METHOD FOR OBJECT REMOVAL BASED ON PATCH LOCAL FEATURE AND SPARSE REPRESENTATION

机译：基于斑块局部特征和稀疏表示的物体去除新方法

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The traditional inpainting methods for object removal from image need to traverse the entire source region and search for the exemplar patches, so it may take a long time to recover the image. Furthermore, these methods simply use the Sum of Squared Differences (SSD) to measure the degree of similarity, which may result in that the target patch is replaced by the inappropriate exemplar patch, and thus there is a risk of introducing undesired objects in recovered image. In view of this situation, we propose a new inpainting method based on patch local feature and sparse representation. First of all, we classify the patches according to their local feature. For smooth patch, we calculate its sparse coefficient over a redundant dictionary, and then recover it using the dictionary and the sparse coefficient. For texture patch, we recover it using the Criminisi method so as to protect the image details. In this way, the sparse representation of the patch can be skillfully used to remove objects from an image. A number of examples on real images show that, the proposed method not only costs less running time, but also avoids introducing undesired objects in recovered images.
机译：用于从图像中去除对象的传统修复方法需要遍历整个源区域并搜索示例补丁，因此恢复图像可能需要很长时间。此外，这些方法仅使用平方差之和（SSD）来测量相似度，这可能导致目标斑块被不合适的示例斑块替代，因此存在在恢复图像中引入不需要的对象的风险。 。针对这种情况，我们提出了一种基于补丁局部特征和稀疏表示的修复方法。首先，我们根据补丁的局部特征对其进行分类。对于平滑补丁，我们在冗余字典上计算其稀疏系数，然后使用字典和稀疏系数对其进行恢复。对于纹理补丁，我们使用Criminisi方法对其进行恢复，以保护图像细节。这样，补丁的稀疏表示可以熟练地用于从图像中删除对象。关于真实图像的大量示例表明，该方法不仅耗费更少的运行时间，而且避免了在恢复的图像中引入不需要的对象。

著录项

  • 来源
  • 作者

    Lei Zhang; Baosheng Kang; Benting Liu; Zhenhua Bao;

  • 作者单位

    School of Information Science and Technology Northwest University No. 1, Xuefu Road, Chang'an District, Xi'an 710127, P. R. China,Public Computer Teaching Department Yuncheng University No. 1155, Fudan West Street, Yuncheng 044000, P. R. China;

    School of Information Science and Technology Northwest University No. 1, Xuefu Road, Chang'an District, Xi'an 710127, P. R. China;

    School of Information Science and Technology Northwest University No. 1, Xuefu Road, Chang'an District, Xi'an 710127, P. R. China;

    School of Information Science and Technology Northwest University No. 1, Xuefu Road, Chang'an District, Xi'an 710127, P. R. China;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    Image inpainting; Object removal; Local feature; Sparse representation;

    机译：图像修复;对象清除;本地特征;稀疏表示;

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号