摘要
采用双门限二值化技术和形态学变换技术对金相图像进行处理,使其晶粒和晶界显得清晰,并且探讨了用Freem an链码编码进行图像压缩的有关技术,并给出了一个适合Freem an链码生成的晶界跟踪算法。提出了填充剔除计数法,用于统计图像中晶粒的颗数,也给出了一个晶粒平均面积和直径的估算公式,为金相的进一步分析打下坚实的基础。
The double thresholds binarization and morphological transform are applied to process the metallographical image. They can classify the grain and the grain boundary. Then the techniques about Freeman chain coding for metal micro-image compression have been discussed, and a grain boundary tracking algorithm also is given. Finally, a filling-elimination method have been produced for counting the number of grains in the micro-image, and also an equation for calculating average area and diameter of grains have been presented. They will bring a stability foundation for the deep metallographical analysis.
出处
《计算机应用与软件》
CSCD
北大核心
2006年第8期100-102,共3页
Computer Applications and Software
关键词
金相图像
图像处理
图像压缩
金相分析
Metallographical image Image processing Image compression Metallographical analysis
