计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (27): 169-174.

• 图形、图像、模式识别 • 上一篇    下一篇

基于优化算法与多尺度差分的图像分割方法

高业文,李柏林,熊  鹰   

  1. 西南交通大学 机械工程学院,成都 610031
  • 出版日期:2012-09-21 发布日期:2012-09-24

Image segmentation method based on optimization algorithm and multi-scale gray difference histogram

GAO Yewen, LI Bolin, XIONG Ying   

  1. School of Mechanical Engineering, Southwest Jiaotong University, Chengdu 610031, China
  • Online:2012-09-21 Published:2012-09-24

摘要: 为了使图像分割算法能够满足实时性要求,针对Otsu法计算量大的问题,将类间方差进行连续域扩展,推导出速度较快的黄金分割法;应用灰度差分直方图进行分割,能够根据多尺度灰度差分直方图得到候选集,将搜索次数减少到候选集中元素个数,计算量小、速度快;结合两者可以实现图像的快速分割。仿真实验和实际应用表明,该方法不仅分割效率高,而且能够得到很好的分割效果。

关键词: 图像分割, Otsu法, 类间方差, 黄金分割法, 多尺度, 灰度差分直方图

Abstract: For the image segmentation algorithms to satisfy the requirement of real-time, this paper extends the between-cluster variance in continuous domain considering the vast computation of Otsu method. The golden section method is derived. This paper proposes the segmentation algorithm based on multi-scale gray difference histogram. The algorithm analyzes the multi-scale gray difference histogram and gets the candidate sets. Combination of the two methods then a fast image segmentation algorithm can be achieved. Simulation results and practical applications show that this algorithm is very well in improving the speed, and good segmentation results can be received.

Key words: image segmentation, Otsu method, between-cluster variance, golden section method, multi-scale, gray difference histogram