作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2014, Vol. 40 ›› Issue (12): 235-241. doi: 10.3969/j.issn.1000-3428.2014.12.044

• 图形图像处理 • 上一篇    下一篇

面向图像局部重建的梯度分析

叶学义,惠舒芸,齐珍珍,李汪兵   

  1. 杭州电子科技大学模式识别与信息安全实验室,杭州 310018
  • 收稿日期:2013-12-23 修回日期:2014-02-26 出版日期:2014-12-15 发布日期:2015-01-16
  • 作者简介:叶学义(1973-),男,副教授、博士,主研方向:图像处理,模式识别,信息隐藏;惠舒芸、齐珍珍、李汪兵,硕士。
  • 基金资助:
    国家自然科学基金青年基金资助项目(60802047,60702018);浙江省科技计划基金资助重点项目(2008C21092);浙江省自然科学基金资助项目(R1090138)

Gradient Analysis for Image Local Reconstruction

YE Xueyi,HUI Shuyun,QI Zhenzhen,LI Wangbing   

  1. Lab of Pattern Recognition & Information Security,Hangzhou Dianzi University,Hangzhou 310018,China
  • Received:2013-12-23 Revised:2014-02-26 Online:2014-12-15 Published:2015-01-16

摘要: 在基于场方向导数的图像局部区域重建过程中,梯度计算对图像边缘的重建效果有很大影响。分析选择单向计算梯度分量的必要性,提出梯度计算的就近原则,对参与梯度分量单向计算的相邻点设置优先级,将缺损点与已知点之间的边缘特征反映在已知点的梯度中,进而影响重建过程,从而实现对图像中边缘特征的重建。实验结果表明,采用该梯度计算模式能有效改善对图像边缘特征的重建效果。

关键词: 点源影响函数, 方向导数, 图像局部重建, 梯度分量单向计算, 就近原则

Abstract: In the reconstructing image local region based on directional derivative,gradient calculation plays an important role in improving the effect of the edges’ reconstruction.Gradient component unidirection calculation is chosen,and adjacency principle is proposed to choose an adjacent point between the two to calculate the gradient component.The edges between the lost points and the known can be reflected by the known gradient in time,and it can be reconstructed accurately.Experimental results show that using the gradient calculation model can effectively improve the reconstruction effect of image edge character.

Key words: point-source influence function, directional derivative, image local reconstruction, gradient component unidirection calculation, adjacency principle

中图分类号: