计算机科学 ›› 2012, Vol. 39 ›› Issue (Z11): 359-361.

• 图形图像 • 上一篇    下一篇

基于差分盒维数的空间目标图像分割算法

姚 远,粱志毅   

  1. (西北工业大学航天学院 西安710072)
  • 出版日期:2018-11-16 发布日期:2018-11-16

Image Segmentation for Space Target Besed on Differential Box Counting

  • Online:2018-11-16 Published:2018-11-16

摘要: 基于差分盒子维数提出了一种针对空间目标的图像分割算法。首先根据空间环境目标自然背景与空间目标人造结构的特点差异,从分形理论的相似性上对星空背景进行分析,利用像素部域灰度方法得到目标和背景的边界。其次,在对给定阂值及该阂值下图像的差分盒维数关系进行分析的基础上,提出基于灰度方差的阂值选择方法。最后给出空间目标图像分割算法的流程,通过诸多仿真空间图像处理验证该分割算法是有效的。

关键词: 差分盒维数,图像分割,空间目标

Abstract: This report proposes a new image segmentation method facing space target based on differential box counting(DBC). Firstly,according to features of the natural background of space target, utilizing similarity of fractal theory to analyze nebula background, using pixel region gray variance to obtain the edge between target and background. Secondly, through analyze relation of threshold and the DISC dimension under the threshold, threshold selection method of gray variance is given. Finally, following chart of the image segmentation is given; besides the experiment result also proves that the segmentation can achieve excellent effect.

Key words: Differential box counting, Image segmentation, Space target

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!