节点文献

图像分割技术在金相分析中的应用

IMAGE SEGMENTATION TECHNIQUE APPLIED IN METALLOGRAPHICAL ANALYSIS

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 汤力琨罗代升王正勇龙建忠

【Author】 TANG Li-kun, LUO Dai-sheng, WANG Zheng-yong, LONG Jian-zhong (School of Electronics and Information Engineering, Sichuan Univ., Chengdu 610065, China)

【机构】 四川大学电子信息学院四川大学电子信息学院 成都610065成都610065成都610065

【摘要】 介绍了常用的图像分割方法及其优缺点和金相图像分析中的图像分割方法。针对金相图像分割中存在的问题,提出了两步分割的图像分割方法。这种方法是先用阈值分割法把目标从背景中分割开来,然后采用数学形态学、打孔、找凹点和连分割线的方法进行粘连分割。与已有的金相图像分割方法相比,这种方法具有较强的强健性、自适应性和非监督性。

【Abstract】 In this paper, the traditional methods of image segmentation and their advantages and disadvantages are introduced first. Then the methods of image segmentation for metallographic analysis are introduced. To solve the problems existing in metallographic analysis, a two-step method of image segmentation is proposed. In the method, the objects are segmented first by thresholding. Then the joint or overlapped objects are separated by mathematic morphology, hole opening, concave point finding, and cutting line determination. Compared with the existing segmentation methods in metallographic analysis, this method is robust, adaptive, and unsupervised.

  • 【文献出处】 理化检验(物理分册) ,Physical Testing and Chemical Analysis Part A Physical Testing , 编辑部邮箱 ,2005年05期
  • 【分类号】TG115.21
  • 【被引频次】2
  • 【下载频次】214
节点文献中: 

本文链接的文献网络图示:

本文的引文网络