计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (11): 87-89.

• 学术探讨 • 上一篇    下一篇

量子衍生形态学图像边缘检测方法

谢可夫 罗安 周心一   

  1. 湖南师范大学图像识别与计算机视觉研究所 湖南大学电气与信息工程学院 湖南师范大学图像识别与计算机视觉研究所
  • 收稿日期:2006-11-01 修回日期:1900-01-01 出版日期:2007-04-11 发布日期:2007-04-11
  • 通讯作者: 谢可夫

A Morphological Method Inspired by Quantum for Edge detection of image

KeFu Xie Luo an Zhou xinyi   

  • Received:2006-11-01 Revised:1900-01-01 Online:2007-04-11 Published:2007-04-11
  • Contact: KeFu Xie

摘要: 提出了一种基于量子测量和坍缩的自适应形态运算。该形态运算的结构元素处于叠加态,在结构元素平移的过程中,由平移点邻域的图像局部特征生成测量算子并对该结构元素进行测量使其坍缩到相应的基态—传统结构元素。由于不同像素点的邻域特征不同,从而形成自适应形态运算。一种基于均方差的测量算子生成方法被提出。计算机仿真实验表明由该形态运算所构成的图像边缘检测算子能从噪声干扰的图像中有效地检测出图像边缘。

关键词: 边缘检测, 数学形态学, 量子信息, 量子衍生

Abstract: A morphological operation based on quantum measurement and collapsing is proposed in this paper. The structure element (SE) for this operation is a multi-state structure element similar to superposition of states of quantum mechanics. In process of the morphological operation, a measurement operator can be constructed according to character of pixels in neighborhood of SE’s translating point. The state of this SE can be collapsed to a corresponding basic state , a traditional SE, after being acted by the measurement operator. The form and scale of the collapsed state can be changed with translating point because of different character of that point’s neighborhood and a self-adapted morphological operation can be formed. A method for producing the measurement operator based on mean square error is proposed and it is shown that the morphological gradient constructed by this morphological operation can detect edge of the image corrupted by noise more efficiently than that traditional morphological gradient through computer simulation.

Key words: Edge detection, Mathematics morpgology, Quantum information, Quantum inspired