| 金属图象分割的研究 |
| |
| 引用本文： | 曹力,史忠科.金属图象分割的研究[J].航空材料学报,2006,26(1):11-15. |
| |
| 作者姓名： | 曹力 史忠科 |
| |
| 作者单位： | 1. 南京航空航天大学,民航学院,南京,210016;西北工业大学自动控制系,西安,710072
2. 西北工业大学自动控制系,西安,710072 |
| |
| 基金项目： | 国家高技术研究发展计划(863计划) |
| |
| 摘 要： | 结合金属图象的特点,利用传统阚值分割技术和基于最大熵原则的图象分割技术对金属图象分割进行了分析;并根据金属图象分析要求,提出了先二值化金属图象再进行边缘提取的方法.对比实验结果表明,基于最大熵原则的图象分割技术可以获得较好的分割效果,所提出的方法能够获得连续闭合的晶粒边缘,为下一步的金属图象分析提供了可靠的依据.
|
| 关 键 词： | 金属图象 阚值分割 边缘提取 最大熵 |
| 文章编号： | 1005-5053(2006)01-0011-05 |
| 修稿时间： | 2004年8月12日 |
Studying to Metal Image Segmentation |
| |
| CAO Li,SHI Zhong-ke.Studying to Metal Image Segmentation[J].Journal of Aeronautical Materials,2006,26(1):11-15. |
| |
| Authors: | CAO Li SHI Zhong-ke |
| |
| Abstract: | With the requirement of metal tissue analysis increasing,metal image processing has become an important field of image processing.Combining with some characters of metal image,metal image segementation was analyzed by traditional methods and a method based on maximum entropy principle.Furthermore,a new method of metal image edge detection was presented by binary image.The results of contrastive experiments indicated that the method based on maximum entropy principle could obtain better metal binary images and the edge detection method in the paper could get continuous and closed grain boundaries.These boundaries can provide reliable foundations to next metal image processing. |
| |
| Keywords: | metal image thresholding edge detection maximum entropy |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
